Abstract:Self-supervised learning (SSL) has recently emerged as a powerful approach to learning representations from large-scale unlabeled data, showing promising results in time series analysis. The self-supervised representation learning can be categorized into two mainstream: contrastive and generative. In this paper, we will present a comprehensive comparative study between contrastive and generative methods in time series. We first introduce the basic frameworks for contrastive and generative SSL, respectively, and discuss how to obtain the supervision signal that guides the model optimization. We then implement classical algorithms (SimCLR vs. MAE) for each type and conduct a comparative analysis in fair settings. Our results provide insights into the strengths and weaknesses of each approach and offer practical recommendations for choosing suitable SSL methods. We also discuss the implications of our findings for the broader field of representation learning and propose future research directions. All the code and data are released at \url{

What problem does this paper attempt to address?

### What problems does this paper attempt to solve? This paper aims to solve the following problems by comparing the performance of two self - supervised learning (SSL) methods, contrastive learning and generative learning, in time - series analysis: 1. **Label scarcity problem**: Time - series data labeling is expensive and depends on domain experts, resulting in scarce labeled data. Self - supervised learning, as a promising method for training models using unlabeled data, can effectively address this challenge. 2. **Understanding the advantages and disadvantages of the two SSL methods**: Although both contrastive learning and generative learning have made significant progress in time - series analysis, the understanding of their relative strengths and weaknesses is still limited. This paper provides clear comparison results and application guidelines by comprehensively comparing these two methods. 3. **Selecting appropriate SSL methods**: Through experimental verification, provide practical suggestions for selecting appropriate self - supervised learning methods for different application scenarios, thereby improving the efficiency of model deployment. 4. **Exploring future research directions**: By revealing the limitations of the two methods, point out the direction for future research and promote the application and development of self - supervised learning in a wider range of fields. ### Specific objectives of the paper - **Introduction of the contrastive framework**: Introduce the basic frameworks of contrastive learning and generative learning, with a focus on discussing how to generate supervised signals to guide model optimization. - **Implementation and comparison of classic algorithms**: Implement SimCLR (a representative of contrastive learning) and MAE (a representative of generative learning) respectively, and conduct a comparative analysis under a fair setting. - **Experimental evaluation**: Use the Human Activity Recognition (HAR) dataset and the ECG dataset to conduct experiments and evaluate the performance of the two methods under different label ratios. - **Analysis of resource consumption**: Analyze the differences between the two methods in training time and computational resources. - **Provide practical application suggestions**: Based on the experimental results, give practical suggestions for choosing contrastive learning or generative learning on different datasets. ### Conclusion Through an in - depth comparison of the two self - supervised learning methods, this paper not only fills an important research gap in the field of time - series analysis, but also provides valuable references for researchers and practitioners, helping them make more informed choices in practical applications.

Self-Supervised Learning for Time Series: Contrastive or Generative?

Self-supervised Learning on Graphs: Contrastive, Generative,or Predictive

Self-Supervised Learning for Time Series Analysis: Taxonomy, Progress, and Prospects

Self-supervised Learning: Generative or Contrastive

Towards Unsupervised Time Series Representation Learning: A Decomposition Perspective

Multi-view Self-Supervised Contrastive Learning for Multivariate Time Series

DCS: Debiased Contrastive Learning with Weak Supervision for Time Series Classification

Skip-Step Contrastive Predictive Coding for Time Series Anomaly Detection.

Contrastive learning based self-supervised time-series analysis

Computer Vision Self-supervised Learning Methods on Time Series

Rethinking self-supervised learning for time series forecasting: A temporal perspective

Self-supervised on Graphs: Contrastive, Generative, or Predictive

A Probabilistic Model Behind Self-Supervised Learning

Self-Supervised Contrastive Representation Learning for Semi-Supervised Time-Series Classification

What Constitutes Good Contrastive Learning in Time-Series Forecasting?

A Survey on Self-supervised Learning: Algorithms, Applications, and Future Trends

Understanding self-supervised Learning Dynamics without Contrastive Pairs

Self-supervised Representation Learning for Time Series Via Temporal Contrasting and Transformation

Denoising-Aware Contrastive Learning for Noisy Time Series