Transparent Networks for Multivariate Time Series
This addresses the need for interpretable models in high-stakes domains where time series data is common, offering a transparent alternative to black-box methods.
The paper tackles the lack of transparent models for multivariate time series by proposing GATSM, a novel transparent neural network that captures temporal patterns and handles dynamic-length series, achieving comparable performance to black-box models like RNNs and Transformers.
Transparent models, which are machine learning models that produce inherently interpretable predictions, are receiving significant attention in high-stakes domains. However, despite much real-world data being collected as time series, there is a lack of studies on transparent time series models. To address this gap, we propose a novel transparent neural network model for time series called Generalized Additive Time Series Model (GATSM). GATSM consists of two parts: 1) independent feature networks to learn feature representations, and 2) a transparent temporal module to learn temporal patterns across different time steps using the feature representations. This structure allows GATSM to effectively capture temporal patterns and handle dynamic-length time series while preserving transparency. Empirical experiments show that GATSM significantly outperforms existing generalized additive models and achieves comparable performance to black-box time series models, such as recurrent neural networks and Transformer. In addition, we demonstrate that GATSM finds interesting patterns in time series. The source code is available at https://github.com/gim4855744/GATSM.