Confidence-Guided Learning Process for Continuous Classification of Time Series

TL;DR

This paper introduces a confidence-guided learning method for continuous classification of time series, enabling models to classify data at every time point despite evolving data distributions, inspired by human confidence dynamics.

Contribution

It proposes a novel confidence-guided approach for continuous time series classification that mimics human confidence patterns to improve learning across multiple data distributions.

Findings

C3TS outperforms all baselines in accuracy on four real-world datasets.

The method effectively manages multiple data distributions during learning.

Confidence-guided scheduling enhances model performance in dynamic environments.

Abstract

In the real world, the class of a time series is usually labeled at the final time, but many applications require to classify time series at every time point. e.g. the outcome of a critical patient is only determined at the end, but he should be diagnosed at all times for timely treatment. Thus, we propose a new concept: Continuous Classification of Time Series (CCTS). It requires the model to learn data in different time stages. But the time series evolves dynamically, leading to different data distributions. When a model learns multi-distribution, it always forgets or overfits. We suggest that meaningful learning scheduling is potential due to an interesting observation: Measured by confidence, the process of model learning multiple distributions is similar to the process of human learning multiple knowledge. Thus, we propose a novel Confidence-guided method for CCTS (C3TS). It can imitate the alternating human confidence described by the Dunning-Kruger Effect. We define the objective- confidence to arrange data, and the self-confidence to control the learning duration. Experiments on four real-world datasets show that C3TS is more accurate than all baselines for CCTS.