Big Data has various formats, substandard data models, and diverse semantics. Understanding this vast, continuous, and constantly evolving data is challenging and requires sophisticated data science pipelines that not only turn the raw data into contextual knowledge and insights but also generate meaningful and personalized narratives to facilitate the understanding and interpretation of the Big Data. Existing methods often fail to provide the necessary context and personalization to make data actionable to diverse stakeholders. To achieve this goal, we introduce the Open Story Model (OSM), a novel data science pipeline that: i) transforms the unprocessed data into a structured Knowledge Graph based on a unified graph data model of typed nodes and edges; ii) generates meaningful narratives on top of the Knowledge Graph; and (iii) link narratives to construct personalized stories for various stakeholders based on their goals. We model each story as an interactive data product to allow stakeholders to engage with and manipulate data in real-time to explore, analyze, and visualize the information in ways that suit their specific needs and questions. To demonstrate the practical application and efficacy of the OSM framework, we focus on a use case in the health domain and explore how OSM facilitates the detection and prediction of cancer by transforming the data into personalized stories for various stakeholders.

Automatic image annotation (AIA) is a fundamental and challenging task in computer vision. Considering the correlations between tags can lead to more accurate image understanding, benefiting various applications, including image retrieval and visual search. While many attempts have been made to incorporate tag correlations in annotation models, the method of constructing a knowledge graph based on external knowledge sources and hyperbolic space has not been explored. In this paper, we create an attributed knowledge graph based on vocabulary, integrate external knowledge sources such as WordNet, and utilize hyperbolic word embeddings for the tag representations. These embeddings provide a sophisticated tag representation that captures hierarchical and complex correlations more effectively, enhancing the image annotation results. In addition, leveraging external knowledge sources enhances contextuality and significantly enriches existing AIA datasets. We exploit two deep learning-based models, the Relational Graph Convolutional Network (R-GCN) and the Vision Transformer (ViT), to extract the input features. We apply two R-GCN operations to obtain word descriptors and fuse them with the extracted visual features. We evaluate the proposed approach using three public benchmark datasets. Our experimental results demonstrate that the proposed architecture achieves state-of-the-art performance across most metrics on Corel5k, ESP Game, and IAPRTC-12. • Knowledge graph with hyperbolic embeddings enhances automatic image annotation tags. • External knowledge sources, like WordNet, enrich the annotation datasets. • R-GCN and ViT models effectively fuse textual and visual features. • Calibrated predictions refine class probabilities for more reliable annotations.

Several real-world problems, like molecular biology and chemical reactions, have hidden graphs, and we need to learn the hidden graph using edge-detecting samples. In this problem, the learner receives examples explaining whether a set of vertices induces an edge of the hidden graph. This article examines the learnability of this problem using the PAC and Agnostic PAC learning models. By computing the VC-dimension of hypothesis spaces of hidden graphs, hidden trees, hidden connected graphs, and hidden planar graphs through edge-detecting samples, we also find the sample complexity of learning these spaces. We study the learnability of this space of hidden graphs in two cases, namely for known and unknown vertex sets. We show that the class of hidden graphs is uniformly learnable when the vertex set is known. Furthermore, we prove that the family of hidden graphs is not uniformly learnable but is nonuniformly learnable when the vertex set is unknown.

Incorporating information from other languages can improve the results of tasks in low-resource languages. A powerful method of building functional natural language processing systems for low-resource languages is to combine multilingual pre-trained representations with cross-lingual transfer learning. In general, however, shared representations are learned separately, either across tasks or across languages. This paper proposes a meta-learning approach for inferring natural language in Persian. Alternately, meta-learning uses different task information (such as QA in Persian) or other language information (such as natural language inference in English). Also, we investigate the role of task augmentation strategy for forming additional high-quality tasks. We evaluate the proposed method using four languages and an auxiliary task. Compared to the baseline approach, the proposed model consistently outperforms it, improving accuracy by roughly six percent. We also examine the effect of finding appropriate initial parameters using zero-shot evaluation and CCA similarity.

Autoregressive models are among the most successful neural network methods for estimating a distribution from a set of samples. However, these models, such as other neural methods, need large data sets to provide good estimations. We believe that knowing structural information about the data can improve their performance on small data sets. Masked autoencoder for distribution estimation (MADE) is a well-structured density estimator, which alters a simple autoencoder by setting a set of masks on its connections to satisfy the autoregressive condition. Nevertheless, this model does not benefit from extra information that we might know about the structure of the data. This information can especially be advantageous in case of training on small data sets. In this article, we propose two autoencoders for estimating the density of a small set of observations, where the data have a known Markov random field (MRF) structure. These methods modify the masking process of MADE, according to conditional dependencies inferred from the MRF structure, to reduce either the model complexity or the problem complexity. We compare the proposed methods with some related binary, discrete, and continuous density estimators on MNIST, binarized MNIST, OCR-letters, and two synthetic data sets.

People use social networks to share millions of stories every day, but these stories rarely become viral. Can we estimate the probability that a story becomes a viral cascade? If so, can we find a set of users that are more likely to trigger viral cascades? These estimation and maximization problems are very challenging since both rare-event nature of viral cascades and efficiency requirement should be considered. Unfortunately, this problem still remains largely unexplored to date. In this paper, given temporal dynamics of a network, we first develop an efficient viral cascade probability estimation method, ViCE, that leverages an special importance sampling approximation to achieve high accuracy, even in the cases of very small probability of influence. We then show that the most influential nodes in this model is NP-hard, and develop an efficient viral cascade probability maximization method, ViCEM, that maximizes a surrogate submodular function using a greedy algorithm. Experiments on both synthetic and real-world data show that ViCE can accurately estimate viral cascade probabilities using fewer samples than naive sampling by at least two orders of magnitude, and also ViCEM finds a set of users with higher viral cascade probability than alternatives. Additionally, experiments show that these algorithms are robust across different network topologies.

Concept drift, change in the underlying distribution that data points come from, is an inevitable phenomenon in data streams. Due to increase in the number of data streams' applications such as network intrusion detection, weather forecasting, and detection of unconventional behavior in financial t ransactions; numerous researches have recently been conducted in the area of concept drift detection. An ideal method for concept drift detection should be able to rapidly and correctly identify changes in the underlying distribution of data points and adapt its model as quickly as possible while the memory and processing time is limited. In this paper, we propose a novel explicit method based on ensemble classifiers for detecting concept drift. The method processes samples one by one, and monitors the distribution of ensemble's error in order to detect probable drifts. After detection of a drift, a new classifier will be trained on the new concept in order to keep the model up-to-date. The proposed method has been evaluated on some artificial and real benchmark data sets. The experiments' results show that the proposed method is capable of detecting and adjusting to concept drifts from different types, and it has outperformed well-known state-of-the-art methods. Especially, in the case of high-speed concept drifts.

With the increasing volume of data, the best approach for learning from this data is to exploit an online learning algorithm. Online ensemble methods take advantage of an ensemble of classifiers to predict labels of data. Prediction with expert advice is a well-studied problem in the online ensembl e learning literature. The weighted majority and the randomized weighted majority (RWM) algorithms are two well-known solutions to this problem, aiming to converge to the best expert. Since among some expert, the best one does not necessarily have the minimum error in all regions of data space, defining specific regions and converging to the best expert in each of these regions will lead to a better result. In this paper, we aim to resolve this problem by proposing a novel online ensemble algorithm to the problem of prediction with expert advice. We propose a cascading version of RWM to achieve not only better experimental results but also a better error bound for sufficiently large datasets.

Incremental learning is a learning algorithm that can get new information from new training sets without forgetting the acquired knowledge from the previously used training sets. In this paper, an incremental learning algorithm based on ensemble learning is proposed. Then, an application of the pro posed algorithm for spam filtering is discussed. The proposed algorithm called incremental RotBoost, assumes the environment is stationary. It trains new weak classifiers for newly arriving data, which are added to the ensemble of classifiers. To evaluate the performance of the proposed algorithm, several computer experiments are conducted. The results of computer experiments show the ability of our proposed algorithm for different tasks in the incremental learning. The results also demonstrate that the proposed algorithm can learn incrementally, and it can learn new classes, as well.

One of the challenging problems in artificial intelligence is credit assignment which simply means distributing the credit among a group, such as a group of agents. We made an attempt to meet this problem with the aid of the reinforcement learning paradigm. In this paper, expertness framework is de fined and applied to the multi-agent credit assignment problem. In the expertness framework, the critic agent, who is responsible for distributing credit among agents, is equipped with learning capability, and the proposed credit assignment solution is based on the critic to learn to assign a proportion of the credit to each agent, and the used proportion should be learned by reinforcement learning. The paper also reports the degree of expertness framework robustness and the amount of performance decline in noisy environments. Experimental results show the superiority of the method over the common methods of credit assignment used in lots of different domains and also show that performance reduction with respect to the quantity of the noise is tolerable and the system ultimately converges to the stable and correct behavior, therefore the agents are still capable of efficiently performing in the noisy environments.

Expert finding problem in bibliographic networks has received increased interest in recent years. This problem concerns finding relevant researchers for a given topic. Motivated by the observation that rarely do all coauthors contribute to a paper equally, in this paper, we propose two discriminative methods for realizing leading authors contributing in a scientific publication. Specifically, we cast the problem of expert finding in a bibliographic network to find leading experts in a research group, which is easier to solve. We recognize three feature groups that can discriminate relevant experts from other authors of a document. Experimental results on a real dataset, and a synthetic one that is gathered from a Microsoft academic search engine, show that the proposed model significantly improves the performance of expert finding in terms of all common information retrieval evaluation metrics.

Expert finding in bibliographic networks has received increased interests in recent years. This task concerns with finding relevant researchers for a given topic. Motivated by the observation that rarely do all coauthors contribute to a paper equally, in this paper, we propose a discriminative method to realize leading authors contributing in a scientific publication. Specifically, we cast the problem of expert finding in a bibliographic network to find leading experts in a research group, which is easier to solve. According to some observations, we recognize three feature groups that can discriminate relevant and irrelevant experts. Experimental results on a real dataset, and an automatically generated one that is gathered from Microsoft academic search show that the proposed model significantly improves the performance of expert finding in terms of all common Information Retrieval evaluation metrics.

Studying biological networks helps to gain a better understanding of cellular behaviors. One of the prominent models to study complex interactions in biological networks is the Nested Effects Model (NEM). Based on the Nested Effects Model, we propose two methods for inferring signaling pathways fro m interventional data. In the first method, we search the space of all feasible solutions with an evolutionary approach to maximize a standard Bayesian score. In the second method, sub-models are constructed with informative features and then combined using an averaging method to make the analysis of larger networks computationally possible. We tested our proposed methods in various noise levels on real and artificial networks with different sizes. The networks constructed by our method have a higher level of accuracy compared to the networks inferred by the triplets method introduced by Markowetz. Moreover, our results show a high level of robustness.

Multiple sequence alignment (MSA) is one of the basic and important problems in molecular biology. MSA can be used for different purposes including finding the conserved motifs and structurally important regions in protein sequences and determine evolutionary distance between sequences. Aligning several sequences cannot be done in polynomial time and therefore heuristic methods such as genetic algorithms can be used to find approximate solutions of MSA problems. Several algorithms based on genetic algorithms have been developed for this problem in recent years. Most of these algorithms use very complicated, problem specific and time consuming mutation operators. In this paper, we propose a new algorithm that uses a new way of population initialization and simple mutation and recombination operators. The strength of the proposed GA is using simple mutation operators and also a special recombination operator that does not have problems of similar recombination operators in other GAs. The experimental results show that the proposed algorithm is capable of finding good MSAs in contrast to existing methods, while it uses simple operators with low computational complexity.

In this paper, we introduce a new hybrid learning algorithm, called DTGP, to construct cost-sensitive classifiers. This algorithm uses a decision tree as its basic classifier and the constructed decision tree will be pruned by a genetic programming algorithm using a fitness function that is sensitive to misclassification costs. The proposed learning algorithm has been examined through six cost-sensitive problems. The experimental results show that the proposed learning algorithm outperforms in comparison to some other known learning algorithms like C4.5 or naïve Bayesian.

The cellular learning automaton (CLA), which is a combination of cellular automaton (CA) and learning automaton (LA), is introduced recently. This model is superior to CA because of its ability to learn and is also superior to single LA because it is a collection of LAs which can interact with each other. The basic idea of CLA is to use LA to adjust the state transition probability of stochastic CA. Recently, various types of CLA such as synchronous, asynchronous, and open CLAs have been introduced. In some applications such as cellular networks, we need to have a model of CLA for which multiple LAs reside in each cell. In this paper, we study a CLA model for which each cell has several LAs. It is shown that, for a class of rules called commutative rules, the CLA model converges to a stable and compatible configuration. Two applications of this new model such as channel assignment in cellular mobile networks and function optimization are also given. For both applications, it has been shown through computer simulations that CLA-based solutions produce better results.

Most email users have encountered with spam problems, which have been addressed as a text classification or categorization problem. In this paper, we propose a novel spam detection method that uses ensemble of classifiers based on clustering and selection techniques. There is diversity in genre of e-mail's content and this method can find different topics in emails by clustering. It first computes disjoint clusters of emails, and then a classifier is trained on each cluster. When new email arrives, its cluster is identified. The classifier of the identified cluster is selected to classify the new email. Our method can extract many kinds of topics in emails. The evaluation shows that the algorithm outperforms majority voting.

A solution to channel assignment problem in cellular networks is self-organizing channel assignment algorithm with distributed control. In this paper, we propose three cellular learning automata based dynamic channel assignment algorithms. In the first two algorithms, no information about the status of channels in the whole network will be used by cells for channel assignment whereas in the third algorithm, the additional information regarding status of channels may be gathered and then used by cells in order to allocate channels. The simulation results show that by using the proposed channel assignment algorithms the micro-cellular network can self-organize itself. The simulation results also show that the additional information used by the third algorithm help the cellular learning automata to find an assignment which results in lower blocking probability of calls for the network.

In this paper, two learning automata based adaptive limited fractional guard channel algorithms for cellular mobile networks are proposed. These algorithms try to minimize the blocking probability of new calls subject to the constraint on the dropping probability of the handoff calls. To evaluate the proposed algorithms, computer simulations are conducted. The simulation results show that the performance of the proposed algorithms are close to the performance of the limited fractional guard channel algorithm for which prior knowledge about traffic parameters are needed. The simulation results also show that the proposed algorithms outperforms the recently introduced dynamic guard channel algorithms.

Cartesian genetic programming (CGP) has an unusual genotype representation which makes it more efficient than Genetic programming (GP) in digital circuit design problem. However, to the best of our knowledge, all methods used in evolutionary design of digital circuits deal with rugged, complex search space, which results in long running time to obtain successful evolution. Therefore, employing a method to guide evolution in these spaces can facilitate achieving more reasonable results. It has been claimed that a two-step evolutionary scenario caused by benefit and cost of learning called Baldwin effect can guide evolution in the biology and artificial life. Therefore, we have been motivated to examine this effect on CGP. We observe using this scenario the success rate and evolution time of CGP improves dramatically especially when size of chromosomes increases.

The difficulty of designing cellular automatons' transition rules to perform a particular problem has severely limited their applications. Using a genetic algorithm to evolve cellular automata for fining these rules, is a good solution. Conventional evolutionary methods use random test configurations for calculating fitness values of each transition rule. In this paper, we use Principal Component Analysis to build better test configurations. By emphasizing on diversity between test instances in a test plan, we can evaluate rules better and faster as well as increase their accuracy. In this paper, we propose two models based on this idea. Experimental results on density classification and synchronization tasks prove that our methods are more efficient than the conventional one.

In this paper the Karnaugh and Quine-McCluskey methods are used for symbolic classification problem, and then these methods are compared with other famous available methods. Because the data in classification problem is very large, some changes should be applied in the original Quine-McCluskey (QMC) algorithm. We proposed a new algorithm that applies the QMC algorithm greedily calling it GQMC. It is surprising that GQMC results are most of the time equal to QMC. GQMC is still very slow classifier and it can be used when the number of attributes of the data is small, and the ratio of training data to the all possible data is high.

In this paper, we first introduce a network of learning automata, which we call it as distributed learning automata and then propose some iterative algorithms for solving stochastic shortest path problem. These algorithms use distributed learning automata to find a policy that determines a path from a source node to a destination node with minimal expected cost (length). In these algorithms, at each stage distributed learning automata determines which edges to be sampled. This sampling method may result in decreasing unnecessary samples and hence decreasing the running time of algorithms. It is shown that the shortest path is found with a probability as close as to unity by proper choice of the parameters of the proposed algorithms.

One popular learning algorithm for feedforward neural networks is the backpropagation (BP) algorithm which includes parameters, learning rate (eta), momentum factor (alpha) and steepness parameter (lambda). The appropriate selections of these parameters have large effects on the convergence of the algorithm. Many techniques that adaptively adjust these parameters have been developed to increase speed of convergence. In this paper, we shall present several classes of learning automata based solutions to the problem of adaptation of BP algorithm parameters. By interconnection of learning automata to the feedforward neural networks, we use learning automata scheme for adjusting the parameters eta, alpha, and lambda based on the observation of random response of the neural networks. One of the important aspects of the proposed schemes is its ability to escape from local minima with high possibility during the training period. The feasibility of proposed methods is shown through simulations on several problems.

Despite of the many successful applications of backpropagation for training multi-layer neural networks, it has many drawbocks. For complex problems it may require a long time to train the networks, and it may not train at all. Long training time can be the result of the non-optimal parameters. It is not easy to choose appropriate value of the parameters for a particular problem. In this paper, by interconnection of fixed structure learning automata (FSLA) to the feedforward neural networks, we apply learning automata (LA) scheme for adjusting these parameters based on the observation of random response of neural networks. The main motivation in using learning automata as an adaptation algorithm is to use its capability of global optimization when dealing with multi-modal surface. The feasibility of proposed method is shown through simulations on three learning problems: exclusive-or, encoding problem, and digit recognition. The simulation results show that the adaptation of these parameters using this method not only increases the convergence rate of learning but it increases the likelihood of escaping from the local minima.