Self-supervised Classification of Weather Systems based on Spatiotemporal Contrastive Learning

Abstract

Correlated time series analysis plays an important role in many real-world industries. Learning an efficient representation of this large-scale data for further downstream tasks is necessary but challenging. In this paper, we propose a time-step-level representation learning framework for individual instances via bootstrapped spatiotemporal representation prediction. We evaluated the effectiveness and flexibility of our representation learning framework on correlated time series forecasting and cold-start transferring the forecasting model to new instances with limited data. A linear regression model trained on top of the learned representations demonstrates our model performs best in most cases. Especially compared to representation learning models, we reduce the RMSE, MAE, and MAPE by 37%, 49%, and 48% on the PeMS-BAY dataset, respectively. Furthermore, in real-world metro passenger flow data, our framework demonstrates the ability to transfer to infer future information of new cold-start instances, with gains of 15%, 19%, and 18%. The source code will be released under the GitHub https://github.com/bonaldli/ Spatiotemporal-TS-Representation-Learning.

Deep Industrial Image Anomaly Detection: A Survey

“…7) Currently, image anomaly detection algorithms can be mainly categorized into two tasks: Industrial image anomaly detection and medical image anomaly detection. Although medical images have more modalities than industrial images [185][186][187][188][189] , the two tasks share many similarities in terms of data and experimental settings. However, few studies have explored how to unify these two tasks.…”

MetaCARD: Meta-Reinforcement Learning with Task Uncertainty Feedback via Decoupled Context-Aware Reward and Dynamics Components

“…For example, Probabilistic Embeddings for Actorcritic meta-RL (PEARL) (Rakelly et al 2019) adopts an amortized variational inference method to learn a probabilistic latent representation of prior experiences. Leveraging from the recent advance of representation learning, (Fu et al 2020;Wang et al 2021;Yuan and Lu 2022) employs contrastive learning to train a compact context encoder so as to capture the distribution of tasks. However, a major limitation of these methods is their entangled latent contexts, making it difficult for the agent to distinguish between changes in dynamics or/and reward functions, thus hampering its adaptability in complex environments.…”