Time Series Classification using the Hidden-Unit Logistic Model

TL;DR

The paper introduces the hidden-unit logistic model for time series classification, leveraging binary hidden units with chain structure to capture complex temporal dependencies, demonstrating superior performance across diverse tasks.

Contribution

It proposes a novel hidden-unit logistic model that models complex decision boundaries and temporal dependencies in time series data, outperforming prior models.

Findings

Strong performance on handwritten character recognition

Effective in speech and facial expression recognition

State-of-the-art facial action unit detection

Abstract

We present a new model for time series classification, called the hidden-unit logistic model, that uses binary stochastic hidden units to model latent structure in the data. The hidden units are connected in a chain structure that models temporal dependencies in the data. Compared to the prior models for time series classification such as the hidden conditional random field, our model can model very complex decision boundaries because the number of latent states grows exponentially with the number of hidden units. We demonstrate the strong performance of our model in experiments on a variety of (computer vision) tasks, including handwritten character recognition, speech recognition, facial expression, and action recognition. We also present a state-of-the-art system for facial action unit detection based on the hidden-unit logistic model.