Recent innovations have made it possible to produce millions of distinct nanoparticles on a chip. These vast volumes of data are impossible to analyze manually, necessitating the development of automated tools. In previous work, we created a binary classification machine learning model to select quality nanoparticle images for downstream analysis. In this work, we show that adding a custom image preprocessing step before model training can produce significantly higher-performing models in a fraction of the time and make the model more robust to different image noise levels and microscope acquisition settings. The proposed image processing pipeline effectively cleans raw nanoparticle images, enhances key features, and allows us to use much lower resolution images and simpler neural network model architectures, resulting in higher performance and significant cost savings. Experiments demonstrate superior performance relative to our baseline, including a 15% improvement in recall and more than a 10% increase in accuracy. Given the high cost of downstream analysis, it is critical to minimize false positives in our application, and our best-performing model obtains a precision of 97.3% and weighted F-score of 95.9% on an unseen test set. Additionally, model training time is reduced from 15.5 hours to 32 seconds. We expect that adopting this pipeline for AI-driven automated nanoparticle characterization will offer a considerable speedup in the laboratory, allowing researchers to rapidly and accurately analyze much greater volumes of data and accelerate materials discovery.

Machine Learning (ML) and Deep Learning (DL) have become increasingly popular in the field of materials science for building property prediction models owing to their ability to efficiently extract and understand data-driven relationships between materials composition, structure, and properties. In general, materials property prediction are regression problems with a vector-based input material representation. While fully connected layers have been widely used in deep neural networks to predict materials properties, simply adding more and more layers to create a deep model often degrades their performance due to the vanishing gradient problem, thereby limiting usage. In this paper, we study and propose architectural principles for building deep regression neural networks comprising fully connected layers with numerical vectors that bypass manual feature engineering. We introduce a novel deep regression neural network with branched residual learning, BRNet, consisting of branching of layers to maximize variation of features learned from the input or previous layer and places skip connections after each layer to minimize the information loss due to vanishing gradient. We perform BRNet model training for inorganic material properties using numerical vectors representing the elemental fractions of the compositions of the respective materials and compare its performance against other traditional ML and DL techniques, including ElemNet and IRNet. Using multiple datasets (such as OQMD, MP, JARVIS) for training and testing, we show that BRNet models are significantly more accurate than the state-of-the-art ML methods and DL models for all data sizes by using only raw elemental fractions as input. We also show that BRNet's branched residual learning requires fewer parameters and leads to better convergence during the training phase than other neural networks, thus resulting in faster model training.

Recently, the dramatic increase of the data amounts drives up the demand for data compression among HPC applications. Although many file systems and I/O middlewares have incorporated compression features, few high-level parallel I/O libraries support data compression due to the challenges of achieving scalable performance on HPC systems. This paper presents the design and implementation of the variable compression feature in the Parallel NetCDF library. Our design employs the same concept of chunking used by the HDF5 library, but we focus on enabling I/O aggregation across multiple requests to address the challenges on performance and scalability. We evaluate our solution using the I/O kernel of real-world scientific applications and analyze the impacts of data compression on parallel I/O performance. Our result suggests that handling multiple requests at once can significantly improve the parallel I/O performance on chunked and compressed data.

Synchronous Stochastic Gradient Descent (SGD) with data parallelism, the most popular parallel training strategy for deep learning, suffers from expensive gradient communications. Local SGD with periodic model averaging is a promising alternative to synchronous SGD. The algorithm allows each worker to locally update its own model, and periodically averages the model parameters across all the workers. While this algorithm enjoys less frequent communications, the convergence rate is strongly affected by the number of workers. In order to scale up the local SGD training without losing accuracy, the number of workers should be sufficiently small so that the model converges reasonably fast. In this paper, we discuss how to exploit the degree of parallelism in local SGD while maintaining model accuracy. Our training strategy employs multiple groups of processes and each group trains a local model based on data parallelism. The local models are periodically averaged across all the groups. Based on this hierarchical parallelism, we design a model averaging algorithm that has a cheaper communication cost than allreduce-based approach. We also propose a practical metric for finding the maximum number of workers that does not cause a significant accuracy loss. Our experimental results demonstrate that our proposed training strategy provides a significantly improved scalability while achieving a comparable model accuracy to synchronous SGD.

The demands of increasingly large scientific application workflows lead to the need for more powerful supercomputers. As the scale of supercomputing systems have grown, the prediction of fault tolerance has become an increasingly critical area of study, since the prediction of system failures can improve performance by saving checkpoints in advance. We propose a real-time failure detection algorithm that adopts an event-based prediction model. The prediction model is a convolutional neural network that utilizes both traditional event attributes and additional spatio-temporal features. We present a case study using our proposed method with six years of reliability, availability, and serviceability event logs recorded by Mira, a Blue Gene/Q supercomputer at Argonne National Laboratory. In the case study, we have shown that our failure prediction model is not limited to predict the occurrence of failures in general. It is capable of accurately detecting specific types of critical failures such as coolant and power problems within reasonable lead time ranges. Our case study shows that the proposed method can achieve a F1score of 0.56 for general failures, 0.97 for coolant failures, and 0.86 for power failures.

Training Convolutional Neural Network (CNN) is a computationally intensive task, requiring efficient parallelization to shorten the execution time. Considering the ever-increasing size of available training data, the parallelization of CNN training becomes more important. Data-parallelism, a popular parallelization strategy that distributes the input data among compute processes, requires the mini-batch size to be sufficiently large to achieve a high degree of parallelism. However, training with large batch size is known to produce a low convergence accuracy. In image restoration problems, for example, the batch size is typically tuned to a small value between 16 ~ 64, making it challenging to scale up the training. In this paper, we propose a parallel CNN training strategy that gradually increases the mini-batch size and learning rate at run-time. While improving the scalability, this strategy also maintains the accuracy close to that of the training with a fixed small batch size. We evaluate the performance of the proposed parallel CNN training algorithm with image regression and classification applications using various models and datasets.

Martensite start temperature (MsT) is an important characteristic of steels, knowledge of which is vital for materials engineers to guide the structural design process of steels. It is defined as the highest temperature at which the austenite phase in steel begins to transform to martensite phase during rapid cooling. Here we describe the development and deployment of predictive models for MsT, given the chemical composition of the material. The data-driven models described here are built on a dataset of about 1000 experimental observations reported in published literature, and the best model developed was found to significantly outperform several existing MsT prediction methods. The data-driven analyses also revealed several interesting insights about the relationship between MsT and the constituent alloying elements of steels. The most accurate predictive model resulting from this work has been deployed in an online web-tool that takes as input the elemental alloying composition of a given steel and predicts its MsT. The online MsT predictor is available at http://info.eecs.northwestern.edu/MsTpredictor.

Additive Manufacturing (AM) is a manufacturing paradigm that builds three-dimensional objects from a computer-aided design model by successively adding material layer by layer. AM has become very popular in the past decade due to its utility for fast prototyping such as 3D printing as well as manufacturing functional parts with complex geometries using processes such as laser metal deposition that would be difficult to create using traditional machining. As the process for creating an intricate part for an expensive metal such as Titanium is prohibitive with respect to cost, computational models are used to simulate the behavior of AM processes before the experimental run. However, as the simulations are computationally costly and time-consuming for predicting multiscale multi-physics phenomena in AM, physics-informed data-driven machine-learning systems for predicting the behavior of AM processes are immensely beneficial. Such models accelerate not only multiscale simulation tools but also empower real-time control systems using in-situ data. In this paper, we design and develop essential components of a scientific framework for developing a data-driven model-based real-time control system. Finite element methods are employed for solving time-dependent heat equations and developing the database. The proposed framework uses extremely randomized trees - an ensemble of bagged decision trees as the regression algorithm iteratively using temperatures of prior voxels and laser information as inputs to predict temperatures of subsequent voxels. The models achieve mean absolute percentage errors below 1% for predicting temperature profiles for AM processes. The code is made available for the research community at https://github.com/paularindam/ml-iter-additive.

Materials discovery is crucial for making scientific advances in many domains. Collections of data from experiments and first-principle computations have spurred interest in applying machine learning methods to create predictive models capable of mapping from composition and crystal structures to materials properties. Generally, these are regression problems with the input being a 1D vector composed of numerical attributes representing the material composition and/or crystal structure. While neural networks consisting of fully connected layers have been applied to such problems, their performance often suffers from the vanishing gradient problem when network depth is increased. Hence, predictive modeling for such tasks has been mainly limited to traditional machine learning techniques such as Random Forest. In this paper, we study and propose design principles for building deep regression networks composed of fully connected layers with numerical vectors as input. We introduce a novel deep regression network with individual residual learning, IRNet, that places shortcut connections after each layer so that each layer learns the residual mapping between its output and input. We use the problem of learning properties of inorganic materials from numerical attributes derived from material composition and/or crystal structure to compare IRNet's performance against that of other machine learning techniques. Using multiple datasets from the Open Quantum Materials Database (OQMD) and Materials Project for training and evaluation, we show that IRNet provides significantly better prediction performance than the state-of-the-art machine learning approaches currently used by domain scientists. We also show that IRNet's use of individual residual learning leads to better convergence during the training phase than when shortcut connections are between multi-layer stacks while maintaining the same number of parameters.

DBSCAN is a well-known clustering algorithm which is based on density and is able to identify arbitrary shaped clusters and eliminate noise data. However, existing parallel implementation strategies based on MPI lack fault tolerance and there is no guarantee that their workload is balanced. Although some of Hadoop-based approaches have been proposed, they do not perform well in terms of scalability since the merge process is not efficient.We propose a scalable parallel DBSCAN algorithm by applying a partitioning strategy. It is implemented in Apache Spark. In order to reduce search time, kdtree is used in our algorithm. To achieve better performance and scalability based on kdtree, we adopt an effective partitioning technique aimed at producing balanced sub-domains which can be computed within Spark executors. Moreover, we came up with a new merging technique: through mapping the relationship between the local points and their bordering neighbors, all the partial clusters which are generated in executors are merged to form the final complete clusters. We have observed and verified (through experiments) that this merging approach is very effective in reducing the time taken for the merge phase and very scalable with increasing the number of processing cores and the generated partial clusters.We implemented the algorithm in Java, evaluated its scalability by using different number of processing cores, and using real and synthetic datasets containing up to several hundred million high-dimensional points. We used three scales of datasets to evaluate our implementation. For small scale, we use 50k, 100k, and 500k data points, obtaining up to a factor of 14.9 speedup when using 16 cores. For medium scale, we use 1.0m, 1.5m, and 1.9m data points, obtaining a factor of 109.2 speedup when using 128 cores. For large scale, we use 61.0m, 91.5m, and 115.9m data points, obtaining a factor of 8344.5 speedup when using 16384 cores.

Intel Xeon Phi is a processor based on MIC architecture that contains a large number of compute cores with a high local memory bandwidth and 512-bit vector processing units. To achieve high performance on Xeon Phi, it is important for programmers to explore all the software features provided by the Intel compiler and libraries to fully utilize the new hardware resources. In this paper, we use the K-Means algorithm to study the performance of various Intel software settings available for Xeon Phi and their impacts to the performance of K-means. At first we examine different memory layouts for storing data points using Intel compiler-intrinsic functions. During distance calculation, the computational kernel of K-means, when the size of individual input data points is not vector-friendly, we pad the data points to align with the VPU width. At last, we implement a parallel reduction to increase memory access parallelism and cache hits. These techniques enable us to successfully take advantage of thread-level parallelism and data-level parallelism on Xeon Phi. Experimental results demonstrate large performance gains over the default auto-vectorization approach. The K-Means implemented with the proposed techniques achieves up to 68.65% and 56.14% performance improvements for aligned datasets and unaligned datasets, respectively. For high-dimensional aligned datasets, we achieved up to 53.49% performance improvement on a large-scale parallel computer.

This paper explores the idea of modeling a large image data collection of polycrystal electron patterns, in order to detect insights in understanding materials discovery. There is an emerging interest in applying big data processing, management and modeling methods to scientific images, which often come in a form and with patterns only interpretable to domain experts. While large-scale machine learning approaches have demonstrated certain superiority in analyzing, summarizing, and providing an understandable route to data types like natural images, speeches and texts, scientific images is still a relatively unexplored area. Deep convolutional neural networks, despite their recent triumph in natural image understanding, are still rarely seen adapted to experimental microscopic images, especially in a large scale. To the best of our knowledge, we present the first deep learning solution towards a scientific image indexing problem using a collection of over 300K microscopic images. The result obtained is 54% better than a dictionary lookup method which is state-of-the-art in the materials science society.

Recent progress in big data and computer vision with deep learning models has gained a lot of attention. Deep learning has been performed on tasks such as image classification, object detection, image segmentation, image captioning, visual question and answering, using large collections of annotated images. This calls for more curated large image datasets with clearer descriptions, cleaner contents, and diversified usability. However, the curation and labeling of such datasets can be labor-intensive. In this paper, we present PinterNet, an algorithm for automatic curation and label generation from noisy textual descriptions, and also publish a big image dataset containing over 110K images automatically labeled with their themes. Our dataset is hierarchical in nature, it has high level category information which we refer as verticals with fine-grained thematic labels at lower level. This advocates a new type of hierarchical theme classification problem closer to human cognition and of business value. We provide benchmark performances using deep learning models based on AlexNet architecture with different pre-training schemes for this novel task and new data.

Fatigue strength is one of the most important mechanical properties of steel. High cost and time for fatigue testing, and potentially disastrous consequences of fatigue failures motivates the development of predictive models for this property. We have developed advanced data-driven ensemble predictive models for this purpose with an extremely high cross-validated accuracy of >98\%, and have deployed these models in a user-friendly online web-tool, which can make very fast predictions of fatigue strength for a given steel represented by its composition and processing information. Such a tool with fast and accurate models is expected to be a very useful resource for the materials science researchers and practitioners to assist in their search for new and improved quality steels. The web-tool is available at http://info.eecs.northwestern.edu/SteelFatigueStrengthPredictor

Geostatistical interpolation is the process that uses existing data and statistical models as inputs to predict data in unobserved spatio-temporal contexts as output. Kriging is a well-known geostatistical interpolation method that minimizes mean square error of prediction. The result interpolated by Kriging is accurate when consistency of statistical properties in data is assumed. However, without this assumption, Kriging interpolation has poor accuracy. To address this problem, this paper presents a new filtering-based clustering algorithm that partitions data into clusters such that the interpolation error within each cluster is significantly reduced, which in turn improves the overall accuracy. Comparisons to traditional Kriging are made with two real-world datasets using two error criteria: normalized mean square error(NMSE) and χ2 test statistics for normalized deviation measurement. Our method has reduced NMSE by more than 50% for both datasets over traditional Kriging. Moreover, χ2 tests have also shown significant improvements of our approach over traditional Kriging.

Allergies are one of the most common chronic diseases worldwide. One in five Americans suffer from either allergy or asthma symptoms. With the prevalence of social media, people sharing experiences and opinions on personal health symptoms and concerns on social media are increasing. Mining those publicly available health related data potentially provides valuable healthcare insights. In this paper, we propose a real-time allergy surveillance system that first classifies tweets to identify those that mention actual allergy incidents using bag-of-words model and NaiveBayesMultinomial classifier and applies in-depth text and spatiotemporal analysis. Our experimental results show that the proposed system can detect predominant allergy types with high precision and that allergy-related tweet volume is highly correlated to the weather data (daily maximum temperature). We believe that this is the first study that examines a large-scale social media stream for in-depth analysis of allergy activities.

Voting on legislative bills to form new laws serves as a key function of most legislature. Predicting the votes of such deliberative bodies leads to better understanding of government policies and generates actionable strategies for social good. In this paper, we present a novel prediction model that maximizes the usage of publicly accessible heterogeneous data, i.e., bill text and lawmakers' profile data, to carry out effective legislative prediction. In particular, we propose to design a probabilistic prediction model which achieves high consistency with past vote records while ensuring the minimum uncertainty of the vote prediction reflecting the firm legal ground often held by the lawmakers. In addition, the proposed legislative prediction model enjoys the following properties: inductive and analytical solution, abilities to deal with the prediction on new bills and new legislators, and robustness to the missing vote issue. We conduct extensive empirical study using the real legislative data and compare with other representative methods in both quantitative political science and data mining communities. The experimental results clearly corroborate that the proposed method provides superior prediction accuracy with visible performance gain.

Massive bipartite graphs are ubiquitous in real world and have important applications in social networks, biological mechanisms, etc. Consider one billion plus people on Facebook making trillions of connections with millions of organizations. Such big social bipartite graphs are often very skewed and unbalanced, on which traditional indexing algorithms do not perform optimally. In this paper, we propose Arowana, a data-driven algorithm for indexing large unbalanced bipartite graphs. Arowana achieves a high-performance efficiency by building an index tree that incorporates the semantic affinity among unbalanced graphs. Arowana uses probabilistic data structures to minimize space overhead and optimize search. In the experiments, we show that Arowana exhibits significant performance improvements and reduces space overhead over traditional indexing techniques.

Discovering communities to understand and model network structures has been a fundamental problem in several fields including social networks, physics, and biology. Many algorithms have been developed for finding the communities. Modularity based technique is fairly new relative to clustering, though it is very popular currently. Although some fast modularity based algorithms exist for detecting communities, the quality of these solutions is limited. At the other extreme, a clique embodies a basic community as it has the greatest possible edge density. However, the requirement that each pair of vertices be connected is too strict. Therefore, techniques to merge partitioned cliques using a hill-climbing greedy algorithm have been studied to form communities. However, the task of finding cliques is computationally expensive. In this paper, we present a new approach for fast and efficient community detection. We propose a clique guided community detection framework that consists of two phases. In the first phase, the framework finds disjoint cliques. In the second phase, the cliques from the first phase are used to guide the merging of individual vertices until a good quality solution is obtained. For the first phase, we develop an algorithm named MaCH (Maximum Clique Heuristic), which is a new approach to compute disjoint cliques using a heuristic-based branch-and-bound technique. We provide experimental results to demonstrate the efficiency of the new algorithm and compare our approach with other previously proposed algorithms.

We address the problem of large scale probabilistic association rule mining and consider the trade-offs between accuracy of the mining results and quest of scalability on modest hardware infrastructure. We demonstrate how extensions and adaptations of research findings can be integrated in an industrial application, and we present the commercially deployed SILVERBACK framework, developed at Voxsup Inc. SILVERBACK tackles the storage efficiency problem by proposing a probabilistic columnar infrastructure and using Bloom filters and reservoir sampling techniques. In addition, a probabilistic pruning technique has been introduced based on Apriori for mining frequent item-sets. The proposed target-driven technique yields a significant reduction on the size of the frequent item-set candidates. We present extensive experimental evaluations which demonstrate the benefits of a context-aware incorporation of infrastructure limitations into corresponding research techniques. The experiments indicate that, when compared to the traditional Hadoop-based approach for improving scalability by adding more hosts, SILVERBACK - which has been commercially deployed and developed at Voxsup Inc. since May 2011 - has much better run-time performance with negligible accuracy sacrifices.

In this paper, we study a challenging problem of inferring individuals' role and statuses in a professional social network, which is of central importance in workforce optimization and human capital management. Realizing the natural setting of social nodes associated with dual view information, i.e., The local node characteristics and the global network influence, we present a novel model that explores graph regularization techniques and integrates such information to achieve improved prediction performance. In particular, our prediction model is built upon the graph transductive learning framework that encodes an uncertainty regularization term in the conventional empirical risk minimization principle. Through taking advantage of the information from both the local profile and the global network characteristics, the final inference of the role or statues achieves minimum an empirical loss on the labeled set, as well as a minimum uncertainty on the unlabeled social nodes. We perform extensive empirical study using real-world data and compare with representative peer approaches. The experimental results on three real social network data sets show that the proposed model greatly outperforms a number of baseline models and is able to effectively infer in a wide range of scenarios.

We consider the problem of active learning when the categories are represented as a tree with leaf nodes as outputs and internal nodes as clusters of the outputs at multiple granularity. Recent work has improved the traditional techniques by moving beyond “flat” structure through incorporation of the label hierarchy into the uncertainty measure. However, these methods have two major limitations when used. First, these methods roughly use the information in the label structure but do not take into account the training samples, which may lead to a sampling bias due to their crude approximation of the class relations. Second, none of these methods can work in a batch mode to reduce the computational time of training. We propose a batch mode active learning scheme that exploits both the hierarchical structure of the labels and the characteristics of the training data to select the most informative data for human labeling. We achieve this goal by first using an approach based on graph embedding that embeds the relationships between the labels and data points in a transformed low-dimensional space. Then, we compute uncertainty by calculating the variance among the points and the labels in the embedding space. Finally, the selection criterion is designed to construct batches and incorporate a diversity measure. Experimental results indicate that our technique achieves a notable improvement in performance over the state-of-the-art approaches.

Community detection in real-world graphs presents a number of challenges. First, even if the number of detected communities grows linearly with the graph size, it becomes impossible to manually inspect each community for value added to the application knowledge base. Mining for communities with query nodes as knowledge priors could allow for filtering out irrelevant information and for enriching end-users knowledge associated with the problem of interest, such as discovery of genes functionally associated with the Alzheimer's (AD) biomarker genes. Second, the data-intensive nature of community enumeration challenges current approaches that often assume that the input graph and the detected communities fit in memory. As computer systems scale, DRAM memory sizes are not expected to increase linearly, while technologies such as SSD memories have the potential to provide much higher capacities at a lower power-cost point, and have a much lower latency than disks. Out-of-core algorithms and/or database-inspired indexing could provide an opportunity for different design optimizations for query-driven community detection algorithms tuned for emerging architectures. Therefore, this work addresses the need for query-driven and memory-efficient community detection. Using maximal cliques as the community definition, due to their high signal-to-noise ratio, we propose and systematically compare two contrasting methods: indexed-based and out-of-core. Both methods improve peak memory efficiency as much as 1000X compared to the state-of-the-art. However, the index-based method, which also has a 10-to-100-fold run time reduction, outperforms the out-of-core algorithm in most cases. The achieved scalability enables the discovery of diseases that are known to be or likely associated with Alzheimer's when the genome-scale network is mined with AD biomarker genes as knowledge priors.

Social media has become a popular platform that connects people who share information, in particular personal opinions. Through such a fast information exchange mechanism, reputation of individuals, consumer products, or business companies can be quickly built up within a social network. Recently, applications mining social network data start emerging to find the communities sharing the same interests for marketing purposes. Knowing the reputation of social network entities, such as celebrities or business companies, can help develop better strategies for election campaigns or new product advertisements. In this paper, we propose a probabilistic graphical model to collectively measure reputations of entities in social networks. By collecting and analyzing large amount of user activities on Facebook, our model can effectively and efficiently rank entities, such as presidential candidates, professional sport teams, musician bands, and companies, based on their social reputation. The proposed model produces results largely consistent with the two publicly available systems - movie ranking in Internet Movie Database and business school ranking by the US news & World Report - with the correlation coefficients of 0.75 and -0.71, respectively.

We analyze lung transplant data from the United Network for Organ Sharing (UNOS) program with the aim of developing accurate risk prediction models for mortality within 1 year of lung transplant using data mining techniques. The data used in this study is de-identified and consists of 62 predictor attributes, and 1-year posttranplant survial outcome for patients who underwent lung transplant between the years 2005 and 2009. Our dataset had 5,319 such patient instances. Several data mining classification techniques were used on this data along with various data mining optimizations and validations to build predictive models for the abovementioned outcome. Prediction results were evaluated using c-statistic metric, and the highest c-statistic obtained was 0.68. Further, we also applied feature selection techniques to reduce the number of attributes in the model from 50 to 8, without any degradation in c-statistic. The final model was also found to outperform logistic regression, which is the most commonly used technique in predictive healthcare informatics. We believe that the resulting predictive model on the reduced dataset can be quite useful to integrate in a risk calculator to aid both physicians and patients in risk assessment.

We analyze the colon cancer data available from the SEER program with the aim of developing accurate survival prediction models for colon cancer. Carefully designed preprocessing steps resulted in removal of several attributes and applying several supervised classification methods. We also adopt synthetic minority over-sampling technique (SMOTE) to balance the survival and non-survival classes we have. In our experiments, ensemble voting of the three of the top performing classifiers was found to result in the best prediction performance in terms of prediction accuracy and area under the ROC curve. We evaluated multiple classification schemes to estimate the risk of mortality after 1 year, 2 years and 5 years of diagnosis, on a subset of 65 attributes after the data clean up process, 13 attribute carefully selected using attribute selection techniques, and SMOTE balanced set of the same 13 attributes, while trying to retain the predictive power of the original set of attributes. Moreover, we demonstrate the importance of balancing the classes of the data set to yield better results.

Recommender systems are vital to the success of online retailers and content providers. One particular challenge in recommender systems is the “cold start” problem. The word “cold” refers to the items that are not yet rated by any user or the users who have not yet rated any items. We propose Elver to recommend and optimize page-interest targeting on Facebook. Existing techniques for cold recommendation mostly rely on content features in the event of lacking user ratings. Since it is very hard to construct universally meaningful features for the millions of Facebook pages, Elver makes minimal assumption of content features. Elver employs iterative matrix completion technology and nonnegative factorization procedure to work with meagre content inklings. Experiments on Facebook data shows the effectiveness of Elver at different levels of sparsity.

Active learning is a promising way to efficiently build up training sets with minimal supervision. Most existing methods consider the learning problem in a pool-based setting. However, in a lot of real-world learning tasks, such as crowdsourcing, the unlabeled samples, arrive sequentially in the form of continuous rapid streams. Thus, preparing a pool of unlabeled data for active learning is impractical. Moreover, performing exhaustive search in a data pool is expensive, and therefore unsuitable for supporting on-the-fly interactive learning in large scale data. In this paper, we present a systematic framework for stream-based multi-class active learning. Following the reinforcement learning framework, we propose a feedback-driven active learning approach by adaptively combining different criteria in a time-varying manner. Our method is able to balance exploration and exploitation during the learning process. Extensive evaluation on various benchmark and real-world datasets demonstrates the superiority of our framework over existing methods.

Name disambiguation is a challenging and important problem in many domains, such as digital libraries, social media management and people search systems. Traditional methods, based on direct assignment using supervised machine learning techniques, seem to be the most effective, but their performances are highly dependent on the amount of training data, while large data annotation can be expensive and time-consuming requiring hours of manual inspection by a domain expert. To efficiently acquire labeled data, we propose a bootstrapping algorithm for the name disambiguation task based on active learning and crowdsourced labeling. We show that the proposed method can leverage the advantages of exploration and exploitation by combining two strategies, thereby improving the overall quality of the training data at minimal expense. The experimental results on two datasets DBLP and ArnetMiner demonstrate the superiority of our framework over existing methods.

The fast development of hospital information systems (HIS) produces a large volume of electronic medical records, which provides a comprehensive source for exploratory analysis and statistics to support clinical decision-making. In this paper, we investigate how to utilize the heterogeneous medical records to aid the clinical treatments of diabetes mellitus. Diabetes mellitus, simply diabetes, is a group of metabolic diseases, which is often accompanied with many complications. We propose a Symptom-Diagnosis-Treatment model to mine the diabetes complication patterns and to unveil the latent association mechanism between treatments and symptoms from large volume of electronic medical records. Furthermore, we study the demographic statistics of patient population w.r.t. complication patterns in real data and observe several interesting phenomena. The discovered complication and treatment patterns can help physicians better understand their specialty and learn previous experiences. Our experiments on a collection of one-year diabetes clinical records from a famous geriatric hospital demonstrate the effectiveness of our approaches.

We investigate sampling techniques in unbalanced heterogeneous bipartite graphs (UHBGs), which have wide applications in real world web-scale social networks. We propose random walked-based link sampling and stratified sampling for UHBGs and show that they have advantages over generic random walk samplers. In addition, each sampler's node degree distribution parameter estimator statistic is analytically derived to be used as a quality indicator. In the experiments, we apply the two sampling techniques, with a baseline node sampling method, to both synthetic and real Facebook data. The experimental results show that random walk-based stratified sampler has significant advantage over node sampler and link sampler on UHBGs.

The various kinds of booming social media not only provide a platform where people can communicate with each other, but also spread useful domain information, such as career and job market information. For example, LinkedIn publishes a large amount of messages either about people who want to seek jobs or companies who want to recruit new members. By collecting information, we can have a better understanding of the job market and provide insights to job-seekers, companies and even decision makers. In this paper, we analyze the job information from the social network point of view. We first collect the job-related information from various social media sources. Then we construct an inter-company job-hopping network, with the vertices denoting companies and the edges denoting flow of personnel between companies. We subsequently employ graphmining techniques to mine influential companies and related company groups based on the job-hopping network model. Demonstration on LinkedIn data shows that our system JobMiner can provide a better understanding of the dynamic processes and a more accurate identification of important entities in the job market.

Social media is producing massive amounts of data on an unprecedented scale. Here people share their experiences and opinions on various topics, including personal health issues, symptoms, treatments, side-effects, and so on. This makes publicly available social media data an invaluable resource for mining interesting and actionable healthcare insights. In this paper, we describe a novel real-time flu and cancer surveillance system that uses spatial, temporal, and text mining on Twitter data. The real-time analysis results are reported visually in terms of US disease surveillance maps, distribution and timelines of disease types, symptoms, and treatments, in addition to overall disease activity timelines on our project website. Our surveillance system can be very useful not only for early prediction of seasonal disease outbreaks such as flu, but also for monitoring distribution of cancer patients with different cancer types and symptoms in each state and the popularity of treatments used. The resulting insights are expected to help facilitate faster response to and preparation for epidemics and also be very useful for both patients and doctors to make more informed decisions.

We investigate a class of emerging online marketing challenges in social networks; macro behavioral targeting (MBT) is introduced as non-personalized broadcasting efforts to massive populations. We propose a new probabilistic graphical model for MBT. Further, a linear-time approximation method is proposed to circumvent an intractable parametric representation of user behaviors. We compare the proposed model with the existing state-of-the-art method on real datasets from social networks. Our model outperforms in all categories by comfortable margins.

Real-world dynamic systems such as physical and atmosphere-ocean systems often exhibit a hierarchical system-subsystem structure. However, the paradigm of making this hierarchical/modular structure and the rich properties they encode a “first-class citizen” of machine learning algorithms is largely absent from the literature. Furthermore, traditional data mining approaches focus on designing new classifiers or ensembles of classifiers, while there is a lack of study on detecting and correcting prediction errors of existing forecasting (or classification) algorithms. In this paper, we propose DETECTOR, a hierarchical method for detecting and correcting forecast errors by employing the whole-part relationships between the target system and non-target systems. Experimental results show that DETECTOR can successfully detect and correct forecasting errors made by state-of-art classifier ensemble techniques and traditional single classifier methods at an average rate of 22%, corresponding to a 11% average forecasting accuracy increase, in seasonal forecasting of hurricanes and landfalling hurricanes in North Atlantic and North African rainfall.

The latent behavior of a physical system that can exhibit extreme events such as hurricanes or rainfalls, is complex. Recently, a very promising means for studying complex systems has emerged through the concept of complex networks. Networks representing relationships between individual objects usually exhibit community dynamics. Conventional community detection methods mainly focus on either mining frequent subgraphs in a network or detecting stable communities in time-varying networks. In this paper, we formulate a novel problem— detection of predictive and phase-biased communities in contrasting groups of networks , and propose an efficient and effective machine learning solution for finding such anomalous communities. We build different groups of networks corresponding to different system’s phases, such as higher or low hurricane activity, discover phase-related system components as seeds to help bound the search space of community generation in each network, and use the proposed contrast-based technique to identify the changing communities across different groups. The detected anomalous communities are hypothesized (1) to play an important role in defining the target system’s state(s) and (2) to improve the predictive skill of the system’s states when used collectively in the ensemble of predictive models. When tested on the two important extreme event problems—identification of tropical cyclone-related and of African Sahel rainfall-related climate indices—our algorithm demonstrated the superior performance in terms of various skill and robustness metrics, including 8–16 % accuracy increase, as well as physical interpretability of detected communities. The experimental results also show the efficiency of our algorithm on synthetic datasets.

Most of the existing active learning algorithms assume all the category labels as independent or consider them in a "flat" structure. However, in reality, there are many applications in which the set of possible labels are often organized in a hierarchical structure. In this paper, we consider the problem of active learning when the categories are represented as a tree. Our goal is to exploit the structure information of the label tree in active learning to select the most informative samples to be labeled. We propose an algorithm that estimates the semantic space, embedding the category hierarchy. In this space, each category label is represented as a prototype and the uncertainty is measured using a variance-based fashion. We also demonstrate notable performance improvement with the proposed approach on synthetic and real datasets.

Social media websites are currently central hubs on the Internet. Major online social media platforms are not only places for individual users to socialize but are increasingly more important as channels for companies to advertise, public figures to engage, etc. In order to optimize such advertising and engaging efforts, there is an emerging challenge for knowledge discovery on today's Internet. The goal of knowledge discovery is to understand the entire online social landscape instead of merely summarizing the statistics. To answer this challenge, we have created VOXSUP as a unified social engagement framework. Unlike most existing tools, VOXSUP not only aggregates and filters social data from the Internet, but also provides what we call Voxsupian Knowledge Discovery (VKD). VKD consists of an almost human-level understanding of social conversations at any level of granularity from a single comment sentiment to multi-lingual inter-platform user demographics. Here we describe the technologies that are crucial to VKD, and subsequently go beyond experimental verification and present case studies from our live VOXSUP system.

This paper proposes a method based on conditional random fields to incorporate sentence structure (syntax and semantics) and context information to identify sentiments of sentences within a document. It also proposes and evaluates two different active learning strategies for labeling sentiment data. The experiments with the proposed approach demonstrate a 5-15% improvement in accuracy on Amazon customer reviews compared to existing supervised learning and rule-based methods.

Processing spatio-temporal queries pertaining to the whereabouts of a large number of mobile entities has traditionally been the topic of the Moving Objects Databases (MOD)research. More recently, due to the advances in sensing and communication technologies, part of the Wireless Sensor Networks(WSN) applications have focused on tracking of mobile objects. These two observations are enough of to warrant a "call" for a confluence of two relatively new but established disciplines. However we observe that a research field of its own right and, historically older than both MOD and WSN - traffic/transportation management - can also capitalize on merging the existing experiences for its own information fusion desiderata In this talk, we will overview applications from seemingly disparate domains and identify their commonalities in terms of the spatio-temporal contexts, and we will discuss how a reactive behavior with pro-active consequences can be efficiently used for large-scale management of mobile data and meta-data.

We address the problem of efficient processing of spatio-temporal range queries for moving objects whose whereabouts in time are not known exactly. The fundamental question tackled by such queries is, given a spatial region and a temporal interval, retrieve the objects that were inside the region during the given interval. As earlier works have demonstrated, when the location, time information is uncertain, syntactic constructs are needed to capture the impact of the uncertainty, along with the corresponding processing algorithms. In this work, we focus on the uncertainty model that represents the whereabouts in-between two known locations as a bead and an uncertain trajectory is represented as a necklace -- a sequence of beads. For each syntactic variant of the range query, we present the respective processing algorithms and, in addition, we propose pruning strategies that speed up the generation of the queries' answers. We also present the experimental observations that quantify the benefits of our proposed methodologies.

Biclustering refers to simultaneous clustering of objects and their features. Use of biclustering is gaining momentum in areas such as text mining, gene expression analysis and collaborative filtering. Due to requirements for high performance in large scale data processing applications such as Collaborative filtering in E-commerce systems and large scale genome-wide gene expression analysis in microarray experiments, a high performance prallel/distributed solution for biclustering problem is highly desirable. Recently, Ahmad et al [1] showed that Bipartite Spectral Partitioning, which is a popular technique for biclustering, can be reformulated as a graph drawing problem where objective is to minimize Hall's energy of the bipartite graph representation of the input data. They showed that optimal solution to this problem is achieved when nodes are placed at the barycenter of their neighbors. In this paper, we provide a parallel algorithm for biclustering based on this formulation. We show that parallel energy minimization using barycenter heuristic is embarrassingly parallel. The challenge is to design a bi-cluster identification algorithm which is scalable as well as accurate. We show that our parallel implementation is not just extremely scalable, it is comparable in accuracy as well with serial implementation. We have evaluated proposed parallel biclustering algorithm with large synthetic data sets on upto 256 processors. Experimental evaluation shows large superlinear speedups, scalability and high level of accuracy.

Existing algorithms for mining frequent patterns are facing challenges to handle databases (a) of increasingly large sizes, (b) consisting of variable-length, irregularly-spaced data, and (c) with mixed or even unknown properties. In this paper, we propose a novel self-adaptive algorithm D-CLUB that thoroughly addresses these issues by progressively clustering the database into condensed association bitmaps, applying a differential technique to digest and remove dense patterns, and then mining the remaining tiny bitmaps directly through fast aggregate bit operations. The bitmaps are well organized into rectangular two-dimensional matrices and adaptively refined in regions that necessitate further computation. We show that this approach not only drastically cuts down the original database size but also largely reduces and simplifies the mining computation for a wide variety of datasets and parameters. We compare D-CLUB with various state-of-the-art algorithms and show significant performance improvement in all cases.

MEMS-based storage systems have recently generated significant interest due to their potential to be faster and more efficient than disks, while providing the non-volatility property. Designing data layouts for these devices is a challenging, important and interesting problem. In this paper, we explore various ways of placing a database on a MEMS-based storage architecture. Three novel data layouts are proposed after considering the MEMS device characteristics and the access patterns arising from queries. We then design the access methodology for each layout and evaluate these layouts based on their respective I/O service times. Overall, our results were able to identify the intricacies of placing data on a MEMS-based storage and also ascertain the large potential of MEMS-based devices for databases.

With the tremendous growth of system memories, memory-resident databases are increasingly becoming important in various domains. Newer memories provide a structured way of storing data in multiple chips, with each chip having a bank of memory modules. Current memory-resident databases are yet to take full advantage of the banked storage system, which offers a lot of room for performance and energy optimizations. In this paper, we identify the implications of a banked memory environment in supporting memory-resident databases, and propose hardware (memory-directed) and software (query-directed) schemes to reduce the energy consumption of queries executed on these databases. Our results show that high-level query-directed schemes (hosted in the query optimizer) better utilize the low-power modes in reducing the energy consumption than the respective hardware schemes (hosted in the memory controller), due to their complete knowledge of query access patterns. We extend this further and propose a query restructuring scheme and a multi-query optimization. Queries are restructured and regrouped based on their table access patterns to maximize the likelihood that data accesses are clustered. This helps increase the inter-access idle times of memory modules, which in turn enables a more effective control of their energy behavior. This heuristic is eventually integrated with our hardware optimizations to achieve maximum savings. Our experimental results show that the memory energy reduces by 90% if query restructuring method is applied along with basic energy optimizations over the unoptimized version. The system-wide performance impact of each scheme is also studied simultaneously.

It is very common that modern large-scale scientific applications employ multiple compute and storage resources in a heterogeneously distributed environment. Working effectively and efficiently in such an environment is one of the major concerns for designing meta-data management systems. The authors present an integrated graphical user interface (GUI) that makes the entire environment virtually an easy-to-use control platform for managing complex programs and their large datasets. To hide the I/O latency when the the user carries out interactive visualization, aggressive prefetching and caching techniques are employed in our GUI. The performance numbers show that the design of our Java GUI has achieved the goals of both high performance and ease-of-use.

Clustering is a key data mining problem. Density and grid based technique is a popular way to mine clusters in a large multi-dimensional space wherein clusters are regarded as dense regions than their surroundings. The attribute values and ranges of these attributes characterize the clusters. Fine grid sizes lead to a huge amount of computation while coarse grid sizes result in loss in quality of clusters found. Also, varied grid sizes result in discovering clusters with different cluster descriptions. The technique of Adaptive grids enables to use grids based on the data distribution and does not require the user to specify any parameters like the grid size or the density thresholds. Further, clusters could be embedded in a subspace of a high dimensional space. We propose a modified bottom-up subspace clustering algorithm to discover clusters in all possible subspaces. Our method scales linearly with the data dimensionality and the size of the data set. Experimental results on a wide variety of synthetic and real data sets demonstrate the effectiveness of Adaptive grids and the effect of the modified subspace clustering algorithm. Our algorithm explores at-least an order of magnitude more number of subspaces than the original algorithm and the use of adaptive grids yields on an average of two orders of magnitude speedup as compared to the method with user specified grid size and threshold.

Decision support systems are important in leveraging information present in data warehouses in businesses like banking, insurance, retail and health care. The multidimensional aspects of a business can be naturally expressed using a multidimensional data model. Data analysis and data mining on these warehouses pose new challenges for traditional database systems. OLAP and data mining operations require summary information on these multidimensional data sets. Query processing for these applications require different views of data for analysis and effective decision making. Data mining techniques can be applied in conjunction with OLAP for an integrated business solution. As data warehouses grow, parallel processing techniques have been applied to enable the use of larger data sets and reduce the time for analysis, thereby enabling evaluation of many more options for decision making. We address: (1) scalability in multidimensional systems for OLAP and multidimensional analysis; (2) integration of data mining with the OLAP framework; and (3) high performance by using parallel processing for OLAP and data mining. We describe our system PARSIMONY-Parallel and Scalable Infrastructure for Multidimensional Online analytical processing. This platform is used both for OLAP and data mining. Sparsity of data sets is handled by using sparse chunks using a bit encoded sparse structure for compression. Techniques for effectively using summary information available in data cubes for data mining are presented for mining association rules and decision tree based classification. These take advantage of the data organization provided by the multidimensional data model. Performance results for high dimensional data sets on a distributed memory parallel machine (IBM SP-2) show good speedup and scalability.

Multidimensional analysis in online analytical processing (OLAP), and scientific and statistical databases (SSDB) use operations requiring summary information on multidimensional data sets. Most common are aggregate operations along one or more dimensions of numerical data values and/or on hierarchies defined on them. Simultaneous calculation of multidimensional aggregates are provided by the Data Cube operator. This is computed only partially if the number of dimensions is large. Queries may either be answered from a materialized cube or calculated on the fly. The multidimensionality of the underlying problem can be represented both in relational and multidimensional databases, the latter being a better fit when query performance is the criteria for judgement. Relational databases are scalable in size for OLAP and multidimensional analysis and efforts are on to make their performance acceptable. On the other hand multidimensional databases provide good performance for such queries, although they are not very scalable. We address scalability in multidimensional systems for analysis in SSDB and OLAP applications. We describe our system PARSIMONY-Parallel and Scalable Infrastructure for Multidimensional Online analytical processing. Sparsity of data sets is handled by using chunks to store data as a sparse set using a bit encoded sparse structure. Chunks provide a multidimensional index structure for efficient dimension oriented data accesses. Operations within and between chunks are a combination of relational and multidimensional operations depending on whether the chunk is sparse or dense. Performance results for high dimensional data sets on a distributed memory parallel machine (IBM SP-2) show good speedup and scalability.

High-performance servers and high-speed networks will form the backbone of the infrastructure required for distributed multimedia information systems. A server for an interactive distributed multimedia system may require thousands of gigabytes of storage space and a high I/O bandwidth. In order to maximize the system utilization, and thus minimize the cost, it is essential that the load be balanced among each of the server's components, viz. the disks, the interconnection network and the scheduler. Many algorithms for maximizing retrieval capacity from the storage system have been proposed in the literature. This paper presents techniques for improving the server capacity by assigning media requests to the nodes of a server so as to balance the load on the interconnection network and the scheduling nodes. Five policies for request assignment-round-robin (RR), minimum link allocation (MLA), minimum contention allocation (MCA), weighted minimum link allocation (WMLA) and weighted minimum contention allocation (WMCA)-are developed. The performance of these policies on a server model developed by the authors (1995) is presented. We also consider the issue of file replication, and develop two schemes for storing the replicas: the parent group-based round-robin placement (PGBRRP) scheme, and the group-wide round-robin placement (GWRRP) scheme. The performance of the request assignment policies in the presence of file replication is presented.

Summary information from data in large databases is used to answer queries in On-Line Analytical Processing (OLAP) systems and to build decision support systems over them. The Data Cube is used to calculate and store summary information on a variety of dimensions, which is computed only partially if the number of dimensions is large. Queries posed on such systems are quite complex and require different views of data. These may either be answered from a materialized cube in the data cube or calculated on the fly. Further, data mining for associations can be performed on the data cube. Analytical models need to capture the multidimensionality of the underlying data, a task for which multidimensional databases are well suited. Multidimensional databases store data in multidimensional structure on which analytical operations are performed. A challenge for these systems is how to handle large data sets in a large number of dimensions. This paper presents a parallel OLAP infrastructure for ...

A performance evaluation of two classes of distributed deadlock detection algorithms, namely, set-based and probe-based distributed deadlock detection algorithms, is presented. The performance evaluation is performed on a simulated distributed database by implementing the algorithms. The performance evaluation shows two main results. First, set-based algorithms outperform probe-based algorithms. Second, current analytical models of distributed deadlock detection are very optimistic because they only compute the overhead of deadlock detection when deadlock exists. It is shown that this overhead cost is only a small portion of the total overall cost, that is, the cost of running the algorithm when deadlock does not exist dominates the cost of the algorithm when deadlock does exist.>