With the increasing concerns about privacy and data regulations, federated learning (FL) has been emerging as a solution to train machine learning models collaboratively with non-exchangeable data from multiple clients. As a result of data locality, data is usually not identically or independently (non-IID) distributed across clients, and the non-IID property has long been the key challenge in FL. Furthermore, in real-world cross-silo scenarios, it is ubiquitous that clients are organizations owning private data from multiple domains internally, which exacerbates the non-IID issue. For example, in healthcare applications, each client (hospital) gathers data from patients with heterogeneous demographics. While previous works have made efforts to address the non-IID challenge across clients by assuming various relations among client-level data distributions and enabling personalized models at the client level, they ignore the internal data heterogeneity within each client or require explicit data domain indicators, which are hardly accessible in real-world data. Here, we propose Sample-Level Prototypical Federated Learning (SL-PFL) to bridge the gap. SL-PFL incorporates prototypical learning under the FL framework and provides a fine-grained personalized model for each data sample instead of learning one uniform model for all samples of each client. Meanwhile, it can be trained using data without ground-truth domain indicators. Experimental results demonstrate that our proposed method with sample-level personalized models outperforms existing FL methods with a global model or client-level personalized models on various real-world regression and classification tasks from weather, computer vision, and healthcare applications.

Counterfactual estimation from observations represents a critical endeavor in numerous application fields, such as healthcare and finance, with the primary challenge being the mitigation of treatment bias. The balancing strategy aimed at reducing covariate disparities between different treatment groups serves as a universal solution. However, when it comes to the time series data, the effectiveness of balancing strategies remains an open question, with a thorough analysis of the robustness and applicability of balancing strategies still lacking. This paper revisits counterfactual estimation in the temporal setting and provides a brief overview of recent advancements in balancing strategies. More importantly, we conduct a critical empirical examination for the effectiveness of the balancing strategies within the realm of temporal counterfactual estimation in various settings on multiple datasets. Our findings could be of significant interest to researchers and practitioners and call for a reexamination of the balancing strategy in time series settings.

Causal analysis for time series data, in particular estimating individualized treatment effect (ITE), is a key task in many real world applications, such as finance, retail, healthcare, etc. Real world time series, i.e., large-scale irregular or sparse and intermittent time series, raise significant challenges to existing work attempting to estimate treatment effects. Specifically, the existence of hidden confounders can lead to biased treatment estimates and complicate the causal inference process. In particular, anomaly hidden confounders which exceed the typical range can lead to high variance estimates. Moreover, in continuous time settings with irregular samples, it is challenging to directly handle the dynamics of causality. In this paper, we leverage recent advances in Lipschitz regularization and neural controlled differential equations (CDE) to develop an effective and scalable solution, namely LipCDE, to address the above challenges. LipCDE can directly model the dynamic causal relationships between historical data and outcomes with irregular samples by considering the boundary of hidden confounders given by Lipschitz constrained neural networks. Furthermore, we conduct extensive experiments on both synthetic and real world datasets to demonstrate the effectiveness and scalability of LipCDE.

A major issue with real-time monitoring is to collect complete data. Hardware or software failures, network issues or, more frequently, time delays can disrupt such a collection. This results in having two versions of the same information: one in real-time but with potentially missing data, and the another, albeit complete, is delayed. Many works have studied how to handle missing data for classification and prediction. However, to the best of our knowledge, they do not consider how to leverage the delayed complete data to assist in learning the representation of real-time available data with missing values. This is despite the fact that the delayed complete data contain all the information (e.g., periodicities and trends). In this paper, we propose a framework to enhance the representation learning of the real-time available data by aligning the representation of past real-time but with missing data to that of past delayed but complete data. We test both a distance metric and contrastive learning to achieve this alignment. We implement our framework on a Transformer-based model and experiment it on three datasets. The efficiency of our solution is evaluated against seven baselines and considering four distinct patterns of missing data. Our experiments show that this proposal has a significant improvement in prediction accuracy (5.21% on average) over the baselines.

Spatiotemporal data aggregated over regions or time windows at various resolutions demonstrate heterogeneous patterns and dynamics in each resolution. Meanwhile, the multi-resolution characteristic provides rich contextual information, which is critical for effective long-sequence forecasting. The importance of such inter-resolution information is more significant in practical cases, where fine-grained data is usually collected via approaches with lower costs but also lower qualities compared to those for coarse-grained data. However, existing works focus on uni-resolution data and cannot be directly applied to fully utilize the aforementioned extra information in multi-resolution data. In this work, we propose Spatiotemporal Koopman Multi-Resolution Network (ST-KMRN), a physics-informed learning framework for long-sequence forecasting from multi-resolution spatiotemporal data. Our method jointly models data aggregated in multiple resolutions and captures the inter-resolution dynamics with the self-attention mechanism. We also propose downsampling and upsampling modules among resolutions to further strengthen the connections among data of multiple resolutions. Moreover, we enhance the modeling of intra-resolution dynamics with physics-informed modules based on Koopman theory. Experimental results demonstrate that our proposed approach achieves the best performance on the long-sequence forecasting tasks compared to baselines without a specific design for multi-resolution data.

Modeling the dynamics of real-world physical systems is critical for spatiotemporal prediction tasks, but challenging when data is limited. The scarcity of real-world data and the difficulty in reproducing the data distribution hinder directly applying meta-learning techniques. Although the knowledge of governing partial differential equations (PDE) of the data can be helpful for the fast adaptation to few observations, it is mostly infeasible to exactly find the equation for observations in real-world physical systems. In this work, we propose a framework, physics-aware meta-learning with auxiliary tasks, whose spatial modules incorporate PDE-independent knowledge and temporal modules utilize the generalized features from the spatial modules to be adapted to the limited data, respectively. The framework is inspired by a local conservation law expressed mathematically as a continuity equation and does not require the exact form of governing equation to model the spatiotemporal observations. The proposed method mitigates the need for a large number of real-world tasks for meta-learning by leveraging spatial information in simulated data to meta-initialize the spatial modules. We apply the proposed framework to both synthetic and real-world spatiotemporal prediction tasks and demonstrate its superior performance with limited observations.

Vast amount of data generated from networks of sensors, wearables, and the Internet of Things (IoT) devices underscores the need for advanced modeling techniques that leverage the spatio-temporal structure of decentralized data due to the need for edge computation and licensing (data access) issues. While federated learning (FL) has emerged as a framework for model training without requiring direct data sharing and exchange, effectively modeling the complex spatio-temporal dependencies to improve forecasting capabilities still remains an open problem. On the other hand, state-of-the-art spatio-temporal forecasting models assume unfettered access to the data, neglecting constraints on data sharing. To bridge this gap, we propose a federated spatio-temporal model -- Cross-Node Federated Graph Neural Network (CNFGNN) -- which explicitly encodes the underlying graph structure using graph neural network (GNN)-based architecture under the constraint of cross-node federated learning, which requires that data in a network of nodes is generated locally on each node and remains decentralized. CNFGNN operates by disentangling the temporal dynamics modeling on devices and spatial dynamics on the server, utilizing alternating optimization to reduce the communication cost, facilitating computations on the edge devices. Experiments on the traffic flow forecasting task show that CNFGNN achieves the best forecasting performance in both transductive and inductive learning settings with no extra computation cost on edge devices, while incurring modest communication cost.

This paper introduces how to infer trust relationships from billion-scale networked data to benefit Alibaba E-Commerce business. To effectively leverage the network correlations between labeled and unlabeled relationships to predict trust relationships, we formalize trust into multiple types and propose a graphical model to incorporate type-based dyadic and triadic correlations, namely eTrust. We also present a fast learning algorithm in order to handle billion-scale networks. Systematically, we evaluate the proposed methods on four different genres of datasets with labeled trust relationships: Alibaba, Epinions, Ciao, and Advogato. Experimental results show that the proposed methods achieve significantly better performance than several comparison methods (+1.7-32.3% by accuracy; p <; <; 0:01, with t-test). Most importantly, when handling the real large networked data with over 1,200,000,000 edges (Ali-large), our method achieves 2,000× speedup to infer trust relationships, comparing with the traditional graph learning algorithms. Finally, we have applied the inferred trust relationships to Alibaba E-commerce platform: Taobao, and achieved 2.75 percent improvement on gross merchandise volume (GMV).

Deep learning models (aka Deep Neural Networks) have revolutionized many fields including computer vision, natural language processing, speech recognition, and is being increasingly used in clinical healthcare applications. However, few works exist which have benchmarked the performance of the deep learning models with respect to the state-of-the-art machine learning models and prognostic scoring systems on publicly available healthcare datasets. In this paper, we present the benchmarking results for several clinical prediction tasks such as mortality prediction, length of stay prediction, and ICD-9 code group prediction using Deep Learning models, ensemble of machine learning models (Super Learner algorithm), SAPS II and SOFA scores. We used the Medical Information Mart for Intensive Care III (MIMIC-III) (v1.4) publicly available dataset, which includes all patients admitted to an ICU at the Beth Israel Deaconess Medical Center from 2001 to 2012, for the benchmarking tasks. Our results show that deep learning models consistently outperform all the other approaches especially when the 'raw' clinical time series data is used as input features to the models.