Accurate traffic prediction is indispensable for intelligent traffic management. The availability of large-scale road sensing data collected by connected wireless sensors and mobile devices have provided unrealized potential for traffic prediction. However, sensory data is often incomplete due to various factors in the process of data acquisition and transmission. The missingness of traffic data brings a key challenge to the traffic prediction task since the state-of-the-art ML-based traffic prediction models (e.g., Graph Convolutional Networks (GCN)) often rely on spatial and temporal completion of the data. Moreover, existing GCN-based methods usually build a static graph based on geographical distances and are limited in their ability to capture the time-evolving relationships amongst road segments. In this paper, we develop a heterogeneous spatio-temporal prediction framework for traffic prediction using incomplete historical data. In the framework, we build multiple graphs to explicitly model the dynamic correlations among road segments from both geographical and historical aspects, and employ recurrent neural networks to capture temporal correlations for each road segment. We impute missing values in a recurrent process, which is seamlessly embedded in the prediction framework so they can be jointly trained. The proposed framework is evaluated on a public dataset of static sensors and a private dataset collected by our roving sensor system. Experimental results show the effectiveness of the proposed framework compared to state-of-the-art methods, and indicate the potential to be deployed into real-world traffic prediction systems.

Spot-level parking availability information (the availability of each spot in a parking lot) is in great demand, as it can help reduce time and energy waste while searching for a parking spot. In this article, we propose a crowdsensing system called SpotE that can provide spot-level availability in a parking lot using drivers’ smartphone sensors. SpotE only requires the sensor data from drivers’ smartphones, which avoids the high cost of installing additional sensors and enables large-scale outdoor deployment. We propose a new model that can use the parking search trajectory and final destination (e.g., an exit of the parking lot) of a single driver in a parking lot to generate the probability profile that contains the probability of each spot being occupied in a parking lot. To deal with conflicting estimation results generated from different drivers, due to the variance in different drivers’ parking behaviors, a novel aggregation approach SpotE-TD is proposed. The proposed aggregation method is based on truth discovery techniques and can handle the variety in Quality of Information of different vehicles. We evaluate our proposed method through a real-life deployment study. Results show that SpotE-TD can efficiently provide spot-level parking availability information with a 20% higher accuracy than the state-of-the-art.

Missing data, which commonly appears in multivariate time series, has been widely recognized as a key challenge in time series analysis. Many commonly used imputation methods either ignore the temporal dependencies of time series data, or do not adequately utilize the relationships among variables. State-ofthe-art methods on time series imputation are built on Recurrent Neural Networks (RNNs), which utilize the historical information to estimate current values sequentially. However, RNNs rely heavily on the output of nearby timestamps, which may lead to important information lost for long sequences. Moreover, individual variables typically present different dynamics and missingness patterns, which is neglected by the global RNN hidden states. In this paper, we propose an imputation framework to learn both global and local dependencies of multivariate time series, as well as a multi-dimensional self-attention to learn capture distant correlations across both time and feature. Extensive experiments show that the proposed framework outperforms the state-of-the-art methods in the imputation task, and benefits the downstream task.

Estimates of road grade/slope can add another dimension of information to existing 2D digital road maps. Integration of road grade information will widen the scope of digital map’s applications, which is primarily used for navigation, by enabling driving safety and efficiency applications such as Advanced Driver Assistance Systems (ADAS), eco-driving, etc. The huge scale and dynamic nature of road networks make sensing road grade a challenging task. Traditional methods oftentimes suffer from limited scalability and update frequency, as well as poor sensing accuracy. To overcome these problems, we propose a cost-effective and scalable road grade estimation framework using sensor data from smartphones. Based on our understanding of the error characteristics of smartphone sensors, we intelligently combine data from accelerometer, gyroscope and vehicle speed data from OBD-II/smartphone’s GPS to estimate road grade. To improve accuracy and robustness of the system, the estimations of road grade from multiple sources/vehicles are crowd-sourced to compensate for the effects of varying quality of sensor data from different sources. Extensive experimental evaluation on a test route of 9km demonstrates the superior performance of our proposed method, achieving 5× improvement on road grade estimation accuracy over baselines, with 90% of errors below 0.3°.

Annotated data samples in real-world applications are often limited. Meta-learning, which utilizes prior knowledge learned from related tasks and generalizes to new tasks of limited supervised experience, is an effective approach for few-shot learning. However, standard meta-learning with globally shared knowledge cannot handle the task heterogeneity problem well, i.e., tasks lie in different distributions. Recent advances have explored several ways to trigger task-dependent initial parameters or metrics, in order to customize task-specific information. These approaches learn task contextual information from data, but ignore external domain knowledge that can help in the learning process. In this paper, we propose a task-adaptive network (TAdaNet) that makes use of a domain-knowledge graph to enrich data representations and provide task-specific customization. Specifically, we learn a task embedding that characterizes task relationships and tailors task-specific parameters, resulting in a task-adaptive metric space for classification. Experimental results on a few-shot image classification problem show the effectiveness of the proposed method. We also apply it on a real-world disease classification problem, and show promising results for clinical decision support.

Utilizing multiple modalities to learn a good distance metric is of vital importance for various clinical applications. However, it is common that modalities are incomplete for some patients due to various technical and practical reasons in healthcare datasets. Existing metric learning methods cannot directly learn the distance metric on such data with missing modalities. Nevertheless, the incomplete data contains valuable information to characterize patient similarity and modality relationships, and they should not be ignored during the learning process. To tackle the aforementioned challenges, we propose a metric learning framework to perform missing modality completion and multi-modal metric learning simultaneously. Employing the generative adversarial networks, we incorporate both complete and incomplete data to learn the mapping relationship between modalities. After completing the missing modalities, we use the nonlinear representations extracted by the discriminator to learn the distance metric among patients. Through jointly training the adversarial generation part and metric learning, the similarity among patients can be learned on data with missing modalities. Experimental results show that the proposed framework learns more accurate distance metric on real-world healthcare datasets with incomplete modalities, comparing with the state-of-the-art approaches. Meanwhile, the quality of the generated modalities can be preserved.

This paper investigates how to estimate instantaneous fuel consumption using smartphones and OBD-II (On-board Diagnostics) adapters. Although most of the new cars have instant miles per gallon readout feature, the readings from those dashboard displays are usually proprietary and are difficult to record. Not to mention older cars do not have this feature. In this paper, we describe a system and associated algorithms to monitor fuel consumption of gasoline-powered vehicles in real time at second-level granularity. Specifically, we propose two algorithms: 1) Powertrain-based Model, which is derived from estimating an engine's fuel injection rate, and 2) Vehicle Dynamics-based Model, which considers fuel consumption in terms of the mechanical work applied to a vehicle. They are designed for vehicles with and without OBD-II adaptors respectively. The proposed system is compatible with most of the passenger vehicles and can be easily deployed. We evaluate our system in a field test and show that it can successfully estimate instantaneous fuel consumption, the average difference between estimation results and the ground truth is about 6%.

A clinically meaningful distance metric, which is learned from measuring patient similarity, plays an important role in clinical decision support applications. Several metric learning approaches have been proposed to measure patient similarity, but they are mostly designed for learning the metric at only one time point/interval. It leads to a problem that those approaches cannot reflect the similarity variations among patients with the progression of diseases. In order to capture similarity information from multiple future time points simultaneously, we formulate a multi-task metric learning approach to identify patient similarity. However, it is challenging to directly apply traditional multi-task metric learning methods to learn such similarities due to the high dimensional, complex and noisy nature of healthcare data. Besides, the disease labels often have clinical relationships, which should not be treated as independent. Unfortunately, traditional formulation of the loss function ignores the degree of labels' similarity. To tackle the aforementioned challenges, we propose mtTSML, a multi-task triplet constrained sparse metric learning method, to monitor the similarity progression of patient pairs. In the proposed model, the distance for each task can be regarded as the combination of a common part and a task-specific one in the transformed low-rank space. We then perform sparse feature selection for each individual task to select the most discriminative information. Moreover, we use triplet constraints to guarantee the margin between similar and less similar pairs according to the ordered information of disease severity levels (i.e. labels). The experimental results on two real-world healthcare datasets show that the proposed multi-task metric learning method significantly outperforms the state-of-the-art baselines, including both single-task and multi-task metric learning methods.

Predicting patients' risk of developing certain diseases is an important research topic in healthcare. Personalized predictive modeling, which focuses on building specific models for individual patients, has shown its advantages on utilizing heterogeneous health data compared to global models trained on the entire population. Personalized predictive models use information from similar patient cohorts, in order to capture the specific characteristics. Accurately identifying and ranking the similarity among patients based on their historical records is a key step in personalized modeling. The electric health records (EHRs), which are irregular sampled and have varied patient visit lengths, cannot be directly used to measure patient similarity due to lack of an appropriate vector representation. In this paper, we build a novel time fusion CNN framework to simultaneously learn patient representations and measure pairwise similarity. Compared to a traditional CNN, our time fusion CNN can learn not only the local temporal relationships but also the contributions from each time interval. Along with the similarity learning process, the output information which is the probability distribution is used to rank similar patients. Utilizing the similarity scores, we perform personalized disease predictions, and compare the effect of different vector representations and similarity learning metrics.

Network Function Virtualization (NFV) provides an effective way to reduce the network provider's cost by allowing multiple Virtual Networks (VNs) to share the underlying physical infrastructure. In the NFV environment, especially when supporting multicast service over the VNs, reliability is a critical requirement in the process of VN mapping since the failure of one virtual node can cause the malfunction of all the subsequent nodes that receive multicasting data from it. In this paper, for the first time, we study how to efficiently map VNs for reliable multicast services, while taking into consideration the max-min fairness of the reliability among distinct VNs. We propose a Mixed Integer Linear Programming (MILP) model to determine the upper bound on the max-min fairness reliability. In addition, an efficient heuristic, namely Uniform Reliability Mutation based Genetic (URMG) algorithm, is developed to address reliable multicast VN mapping with a low computational complexity. By encoding multicast tree construction and link mapping into path selection, taking into consideration the max-min reliability fairness goal, and the networking reliability factors during mutation, URMG can globally optimize the reliability and its fairness of all the multicast VN requests. Through extensive simulations, we demonstrate that URMG achieves close to the optimal reliability fairness with a much lower time complexity than the MILP and yields a significant performance improvement in terms of reliability fairness, bandwidth consumption and transmission delay comparing with other heuristic solutions.

Network Function Virtualization (NFV) allows multiple Virtual Networks (VNs) to share the underlying physical infrastructure via VN mapping, thus improving the utilization of physical resources. In this paper, for the first time, we study the multicast service-oriented VN mapping that can support big data applications over Elastic Optical Networks (EONs). Since the problem of minimizing the spectrum consumption in multicast service-oriented VN mapping is NP-hard, we propose an efficient heuristic algorithm, called Integrated Genetic and Simulated Annealing (IGSA) algorithm to address the problem with low computational complexity. By encoding node mapping, multicast tree construction, link mapping and spectrum requirements in the same gene and auto-adjusted evolution, and utilizing simulated annealing to find the fittest multicast requests mapping order, IGSA can perform joint optimization for all the multicast requests in a global way. Through extensive simulations, we demonstrate that IGSA outperforms the other heuristic solutions in terms of spectrum consumption, blocking probability and normalized throughput, while achieving close to minimum spectrum consumption with a much lower time complexity than MILP.

We design genetic algorithms (GA) to solve the mixed regenerator placement (MRP) problem of lightpaths with different lengths in optical transport networks, and investigate their performance with numerical simulations. By incorporating a theoretical model that can estimate BER changes hop-by-hop along lightpaths, the GA encodes the placements of 1R/2R/3R at intermediate regeneration sites as genes, and takes quality-of-transmission (QoT) and energy-efficiency as the fitness functions. With a relatively small population size (e.g. 30-50), the algorithms obtain multiple qualified MRP results that can satisfy both the QoT and energy requirements within 32 generations, for lightpaths with lengths up to 28 hops. Two crossover and two mutation operators are investigated within the GA. By adjusting the possibilities of the crossover and mutation intelligently, the adaptive scheme outperforms the uniform one by getting a larger percentage of fit individuals. We also propose two selection operators, and demonstrate an adjustable tradeoff between QoT and energy-efficiency.