Search

Enhancing diverse human decision-making processes in an urban environment is a critical issue across various applications, including ride-sharing vehicle dispatching, public transportation management, and autonomous driving. Offline reinforcement learning (RL) is a promising approach to learn and optimize human urban strategies (or policies) from pre-collected human-generated spatial-temporal urban data. However, standard offline RL faces two significant challenges: (1) data scarcity and data heterogeneity, and (2) distributional shift. In this paper, we introduce MODA -- a Multi-Task Offline Reinforcement Learning with Contrastive Data Sharing approach. MODA addresses the challenges of data scarcity and heterogeneity in a multi-task urban setting through Contrastive Data Sharing among tasks. This technique involves extracting latent representations of human behaviors by contrasting positive and negative data pairs. It then shares data presenting similar representations with the target task, facilitating data augmentation for each task. Moreover, MODA develops a novel model-based multi-task offline RL algorithm. This algorithm constructs a robust Markov Decision Process (MDP) by integrating a dynamics model with a Generative Adversarial Network (GAN). Once the robust MDP is established, any online RL or planning algorithm can be applied. Extensive experiments conducted in a real-world multi-task urban setting validate the effectiveness of MODA. The results demonstrate that MODA exhibits significant improvements compared to state-of-the-art baselines, showcasing its capability in advancing urban decision-making processes. We also made our code available to the research community., Comment: KDD 2024

Foundation models, i.e., very large deep learning models, have demonstrated impressive performances in various language and vision tasks that are otherwise difficult to reach using smaller-size models. The major success of GPT-type of language models is particularly exciting and raises expectations on the potential of foundation models in other domains including satellite remote sensing. In this context, great efforts have been made to build foundation models to test their capabilities in broader applications, and examples include Prithvi by NASA-IBM, Segment-Anything-Model, ViT, etc. This leads to an important question: Are foundation models always a suitable choice for different remote sensing tasks, and when or when not? This work aims to enhance the understanding of the status and suitability of foundation models for pixel-level classification using multispectral imagery at moderate resolution, through comparisons with traditional machine learning (ML) and regular-size deep learning models. Interestingly, the results reveal that in many scenarios traditional ML models still have similar or better performance compared to foundation models, especially for tasks where texture is less useful for classification. On the other hand, deep learning models did show more promising results for tasks where labels partially depend on texture (e.g., burn scar), while the difference in performance between foundation models and deep learning models is not obvious. The results conform with our analysis: The suitability of foundation models depend on the alignment between the self-supervised learning tasks and the real downstream tasks, and the typical masked autoencoder paradigm is not necessarily suitable for many remote sensing problems.

When dealing with data from distinct locations, machine learning algorithms tend to demonstrate an implicit preference of some locations over the others, which constitutes biases that sabotage the spatial fairness of the algorithm. This unfairness can easily introduce biases in subsequent decision-making given broad adoptions of learning-based solutions in practice. However, locational biases in AI are largely understudied. To mitigate biases over locations, we propose a locational meta-referee (Meta-Ref) to oversee the few-shot meta-training and meta-testing of a deep neural network. Meta-Ref dynamically adjusts the learning rates for training samples of given locations to advocate a fair performance across locations, through an explicit consideration of locational biases and the characteristics of input data. We present a three-phase training framework to learn both a meta-learning-based predictor and an integrated Meta-Ref that governs the fairness of the model. Once trained with a distribution of spatial tasks, Meta-Ref is applied to samples from new spatial tasks (i.e., regions outside the training area) to promote fairness during the fine-tune step. We carried out experiments with two case studies on crop monitoring and transportation safety, which show Meta-Ref can improve locational fairness while keeping the overall prediction quality at a similar level.

Predicting dissolved oxygen (DO) concentrations in north temperate lakes requires a comprehensive study of phenological patterns across various ecosystems, which highlights the significance of selecting phenological features and feature interactions. Process-based models are limited by partial process knowledge or oversimplified feature representations, while machine learning models face challenges in efficiently selecting relevant feature interactions for different lake types and tasks, especially under the infrequent nature of DO data collection. In this paper, we propose a Nature-Guided Cognitive Evolution (NGCE) strategy, which represents a multi-level fusion of adaptive learning with natural processes. Specifically, we utilize metabolic process-based models to generate simulated DO labels. Using these simulated labels, we implement a multi-population cognitive evolutionary search, where models, mirroring natural organisms, adaptively evolve to select relevant feature interactions within populations for different lake types and tasks. These models are not only capable of undergoing crossover and mutation mechanisms within intra-populations but also, albeit infrequently, engage in inter-population crossover. The second stage involves refining these models by retraining them with real observed labels. We have tested the performance of our NGCE strategy in predicting daily DO concentrations across a wide range of lakes in the Midwest, USA. These lakes, varying in size, depth, and trophic status, represent a broad spectrum of north temperate lakes. Our findings demonstrate that NGCE not only produces accurate predictions with few observed labels but also, through gene maps of models, reveals sophisticated phenological patterns of different lakes.

Fairness-awareness has emerged as an essential building block for the responsible use of artificial intelligence in real applications. In many cases, inequity in performance is due to the change in distribution over different regions. While techniques have been developed to improve the transferability of fairness, a solution to the problem is not always feasible with no samples from the new regions, which is a bottleneck for pure data-driven attempts. Fortunately, physics-based mechanistic models have been studied for many problems with major social impacts. We propose SimFair, a physics-guided fairness-aware learning framework, which bridges the data limitation by integrating physical-rule-based simulation and inverse modeling into the training design. Using temperature prediction as an example, we demonstrate the effectiveness of the proposed SimFair in fairness preservation., Comment: Accepted to AAAI 2024 (preprint)

Modeling environmental ecosystems is critical for the sustainability of our planet, but is extremely challenging due to the complex underlying processes driven by interactions amongst a large number of physical variables. As many variables are difficult to measure at large scales, existing works often utilize a combination of observable features and locally available measurements or modeled values as input to build models for a specific study region and time period. This raises a fundamental question in advancing the modeling of environmental ecosystems: how to build a general framework for modeling the complex relationships amongst various environmental data over space and time? In this paper, we introduce a new framework, FREE, which maps available environmental data into a text space and then converts the traditional predictive modeling task in environmental science to the semantic recognition problem. The proposed FREE framework leverages recent advances in Large Language Models (LLMs) to supplement the original input features with natural language descriptions. This facilitates capturing the data semantics and also allows harnessing the irregularities of input features. When used for long-term prediction, FREE has the flexibility to incorporate newly collected observations to enhance future prediction. The efficacy of FREE is evaluated in the context of two societally important real-world applications, predicting stream water temperature in the Delaware River Basin and predicting annual corn yield in Illinois and Iowa. Beyond the superior predictive performance over multiple baseline methods, FREE is shown to be more data- and computation-efficient as it can be pre-trained on simulated data generated by physics-based models.

Neural networks trained on distilled data often produce over-confident output and require correction by calibration methods. Existing calibration methods such as temperature scaling and mixup work well for networks trained on original large-scale data. However, we find that these methods fail to calibrate networks trained on data distilled from large source datasets. In this paper, we show that distilled data lead to networks that are not calibratable due to (i) a more concentrated distribution of the maximum logits and (ii) the loss of information that is semantically meaningful but unrelated to classification tasks. To address this problem, we propose Masked Temperature Scaling (MTS) and Masked Distillation Training (MDT) which mitigate the limitations of distilled data and achieve better calibration results while maintaining the efficiency of dataset distillation., Comment: ICCV 2023

Human mobility estimation is crucial during the COVID-19 pandemic due to its significant guidance for policymakers to make non-pharmaceutical interventions. While deep learning approaches outperform conventional estimation techniques on tasks with abundant training data, the continuously evolving pandemic poses a significant challenge to solving this problem due to data nonstationarity, limited observations, and complex social contexts. Prior works on mobility estimation either focus on a single city or lack the ability to model the spatio-temporal dependencies across cities and time periods. To address these issues, we make the first attempt to tackle the cross-city human mobility estimation problem through a deep meta-generative framework. We propose a Spatio-Temporal Meta-Generative Adversarial Network (STORM-GAN) model that estimates dynamic human mobility responses under a set of social and policy conditions related to COVID-19. Facilitated by a novel spatio-temporal task-based graph (STTG) embedding, STORM-GAN is capable of learning shared knowledge from a spatio-temporal distribution of estimation tasks and quickly adapting to new cities and time periods with limited training samples. The STTG embedding component is designed to capture the similarities among cities to mitigate cross-task heterogeneity. Experimental results on real-world data show that the proposed approach can greatly improve estimation performance and out-perform baselines., Comment: Accepted at the 22nd IEEE International Conference on Data Mining (ICDM 2022) Full Paper

The eco-toll estimation problem quantifies the expected environmental cost (e.g., energy consumption, exhaust emissions) for a vehicle to travel along a path. This problem is important for societal applications such as eco-routing, which aims to find paths with the lowest exhaust emissions or energy need. The challenges of this problem are three-fold: (1) the dependence of a vehicle's eco-toll on its physical parameters; (2) the lack of access to data with eco-toll information; and (3) the influence of contextual information (i.e. the connections of adjacent segments in the path) on the eco-toll of road segments. Prior work on eco-toll estimation has mostly relied on pure data-driven approaches and has high estimation errors given the limited training data. To address these limitations, we propose a novel Eco-toll estimation Physics-informed Neural Network framework (Eco-PiNN) using three novel ideas, namely, (1) a physics-informed decoder that integrates the physical laws of the vehicle engine into the network, (2) an attention-based contextual information encoder, and (3) a physics-informed regularization to reduce overfitting. Experiments on real-world heavy-duty truck data show that the proposed method can greatly improve the accuracy of eco-toll estimation compared with state-of-the-art methods., Comment: Full version of the paper accepted for the SDM23 conference; Yan Li and Mingzhou Yang contributed equally to this paper

Spatio-temporal machine learning is critically needed for a variety of societal applications, such as agricultural monitoring, hydrological forecast, and traffic management. These applications greatly rely on regional features that characterize spatial and temporal differences. However, spatio-temporal data often exhibit complex patterns and significant data variability across different locations. The labels in many real-world applications can also be limited, which makes it difficult to separately train independent models for different locations. Although meta learning has shown promise in model adaptation with small samples, existing meta learning methods remain limited in handling a large number of heterogeneous tasks, e.g., a large number of locations with varying data patterns. To bridge the gap, we propose task-adaptive formulations and a model-agnostic meta-learning framework that ensembles regionally heterogeneous data into location-sensitive meta tasks. We conduct task adaptation following an easy-to-hard task hierarchy in which different meta models are adapted to tasks of different difficulty levels. One major advantage of our proposed method is that it improves the model adaptation to a large number of heterogeneous tasks. It also enhances the model generalization by automatically adapting the meta model of the corresponding difficulty level to any new tasks. We demonstrate the superiority of our proposed framework over a diverse set of baselines and state-of-the-art meta-learning frameworks. Our extensive experiments on real crop yield data show the effectiveness of the proposed method in handling spatial-related heterogeneous tasks in real societal applications., Comment: In the Thirty-Seventh AAAI Conference on Artificial Intelligence, February 2023

This paper proposes a new data-driven method for predicting water temperature in stream networks with reservoirs. The water flows released from reservoirs greatly affect the water temperature of downstream river segments. However, the information of released water flow is often not available for many reservoirs, which makes it difficult for data-driven models to capture the impact to downstream river segments. In this paper, we first build a state-aware graph model to represent the interactions amongst streams and reservoirs, and then propose a parallel learning structure to extract the reservoir release information and use it to improve the prediction. In particular, for reservoirs with no available release information, we mimic the water managers' release decision process through a pseudo-prospective learning method, which infers the release information from anticipated water temperature dynamics. For reservoirs with the release information, we leverage a physics-based model to simulate the water release temperature and transfer such information to guide the learning process for other reservoirs. The evaluation for the Delaware River Basin shows that the proposed method brings over 10\% accuracy improvement over existing data-driven models for stream temperature prediction when the release data is not available for any reservoirs. The performance is further improved after we incorporate the release data and physical simulations for a subset of reservoirs.

The photogalvanic effect (PGE) occurring in noncentrosymmetric materials enables the generation of a dc photocurrent at zero bias with a high polarization sensitivity, which makes it very attractive in photodetection. However, the magnitude of the PGE photocurrent is usually small, leading to a low photoresponsivity, and therefore hampers its practical application in photodetection. Here, we propose an approach to largely enhancing the PGE photocurrent by applying an inhomogenous mechanical stretch, based on quantum transport simulations. We model a two-dimensional photodetector consisting of the wide-bandgap MgCl$_2$/ZnBr$_2$ vertical van der Waals heterojunction with the noncentrosymmetric $C_{3v}$ symmetry. Polarization-sensitive PGE photocurrent is generated under the vertical illumination of linearly polarized light. By applying inhomogenous mechanical stretch on the lattice, the photocurrent can be largely increased by up to 3 orders of magnitude due to the significantly increased device asymmetry. Our results propose an effective way to enhance the PGE by inhomogenous mechanical strain, showing the potential of the MgCl$_2$/ZnBr$_2$ vertical heterojunction in the low-power UV photodetection., Comment: 6 pages, 6 figures

Mapping of spatial hotspots, i.e., regions with significantly higher rates of generating cases of certain events (e.g., disease or crime cases), is an important task in diverse societal domains, including public health, public safety, transportation, agriculture, environmental science, etc. Clustering techniques required by these domains differ from traditional clustering methods due to the high economic and social costs of spurious results (e.g., false alarms of crime clusters). As a result, statistical rigor is needed explicitly to control the rate of spurious detections. To address this challenge, techniques for statistically-robust clustering (e.g., scan statistics) have been extensively studied by the data mining and statistics communities. In this survey we present an up-to-date and detailed review of the models and algorithms developed by this field. We first present a general taxonomy for statistically-robust clustering, covering key steps of data and statistical modeling, region enumeration and maximization, and significance testing. We further discuss different paradigms and methods within each of the key steps. Finally, we highlight research gaps and potential future directions, which may serve as a stepping stone in generating new ideas and thoughts in this growing field and beyond., Comment: 39 pages, 5 figures, accepted by ACM Computing Surveys (CSUR)

Spatial variability has been observed in many geo-phenomena including climatic zones, USDA plant hardiness zones, and terrestrial habitat types (e.g., forest, grasslands, wetlands, and deserts). However, current deep learning methods follow a spatial-one-size-fits-all(OSFA) approach to train single deep neural network models that do not account for spatial variability. In this work, we propose and investigate a spatial-variability aware deep neural network(SVANN) approach, where distinct deep neural network models are built for each geographic area. We evaluate this approach using aerial imagery from two geographic areas for the task of mapping urban gardens. The experimental results show that SVANN provides better performance than OSFA in terms of precision, recall,and F1-score to identify urban gardens., Comment: Accepted in 1st ACM SIGKDD Workshop on Deep Learning for Spatiotemporal Data, Applications, and Systems (Deepspatial 2020), San Diego, CA, August 24, 2020

Photogalvanic effect (PGE) occurring in noncentrosymmetric materials enables the generation of the open-circuit voltage that is much larger than the bandgap, making it rather attractive in solar cells. However, the magnitude of the PGE photocurrent is usually small, which severely hampers its practical application. Here we propose a mechanism to largely enhance the PGE photocurrent by mechanical strain based on the quantum transport simulations for the two-dimensional nickel-phosphorene-nickel photodetector. Broadband PGE photocurrent governed by the Cs noncentrosymmetry is generated at zero bias under the illumination of linearly polarized light. The photocurrent depends linearly on the device asymmetry, while nonlinearly on the optical absorption. By applying the appropriate mechanical tension stress on the phosphorene, the photocurrent can be substantially enhanced by up to 3 orders of magnitude, which is primarily ascribed to the largely increased device asymmetry. The change in the optical absorption in some cases can also play a critical role in tuning the photocurrent due to the nonlinear dependence. Moreover, the photocurrent can even be further enhanced by the mechanical bending, mainly owing to the considerably enhanced device asymmetry. Our results reveal the dependence of the PGE photocurrent on the device asymmetry and absorption in transport process through a device, and also explore the potentials of the PGE in the self-powered low-dimensional flexible optoelectronics., Comment: 9 pages, 5 figures

Pure spin-current is of central importance in spintronics. Here we propose a two-dimensional (2D) spin-battery system that delivers pure spin-current without an accompanying charge-current to the outside world at zero bias. The principle of the spin-battery roots in the photogalvanic effect (PGE), and the system has good operational stability against structural perturbation, photon energy and other materials detail. The device principle is numerically implemented in the 2D material phosphorene as an example, and first principles calculations give excellent qualitative agreement with the PGE physics. The 2D spin-battery is interesting as it is both a device that generates pure spin-currents, also an energy source that harvests photons. Given the versatile operational space, the spin-battery should be experimentally feasible., Comment: 10 pages, 7 figures

Two-dimensional thermoelectric materials with a figure of merit $ZT$, which is greater than 2.0 at room temperature, would be highly desirable in energy conversion since the efficiency is competitive to conventional energy conversion techniques. Here, we propose that the indium triphosphide (InP$_3$) monolayer offers an extraordinary $ZT$ of 2.2 at 300 K by using quantum calculations within the ballistic thermal transport region. A remarkably low and isotropic phononic thermal conductivity is founded, which is due to flat lattice vibration modes. This low thermal conductivity takes a major responsibility to the impressively high $ZT$. Moreover, a large $ZT$ that is greater than 1.5 can be maintained, even if a 1% mechanic extension is applied on the lattice. These results suggest that the InP$_3$ monolayer is a promising candidate for low dimensional thermoelectric applications.

We report a first-principles study on the tunnel magnetoresistance (TMR) and spin-injection efficiency (SIE) through phosphorene with nickel electrodes under the mechanical tension and bending on the phosphorene region. Both the TMR and SIE are largely improved under these mechanical deformations. For the uniaxial tension ($\varepsilon_y$) varying from 0 to 15\% applied along the armchair transport ({\it y}-)direction of the phosphorene, the TMR ratio is enhanced with a maximum of 107\% at the $\varepsilon_y=10\%$, while the SIE increases monotonously from 8\% up to 43\% with the increasing of the strain. Under the out-of-plane bending, the TMR overall increases from 7\% to 50\% within the bending ratio of 0-3.9\%, and meanwhile the SIE is largely improved to around 70\%, as compared to that (30\%) of the flat phosphorene. Such behaviors of the TMR and SIE are mainly affected by the transmission of spin-up electrons in the parallel configuration, which is highly depended on the applied mechanical tension and bending. Our results indicate that the phosphorene based tunnel junctions have promising applications in flexible electronics., Comment: 5 pages, 6 figures

Monolayer black phosphorus (MBP) is an interesting emerging electronic material with a direct band gap and relatively high carrier mobility. In this work we report a theoretical investigation of nonequilibrium spin injection and spin-polarized quantum transport in MBP from ferromagnetic Ni contacts, in two-dimensional magnetic tunneling structures. We investigate physical properties such as the spin injection efficiency, the tunnel magnetoresistance ratio, spin-polarized currents, charge currents and transmission coefficients as a function of external bias voltage, for two different device contact structures where MBP is contacted by Ni(111) and by Ni(100). While both structures are predicted to give respectable spin-polarized quantum transport, the Ni(100)/MBP/Ni(100) trilayer has the superior properties where the spin injection and magnetoresistance ratio maintains almost a constant value against the bias voltage. The nonequilibrium quantum transport phenomenon is understood by analyzing the transmission spectrum at nonequilibrium., Comment: 6 pages, 6 figures