A Temporal Kolmogorov-Arnold Transformer for Time Series Forecasting

TL;DR

The paper introduces the Temporal Kolmogorov-Arnold Transformer (TKAT), an attention-based model that leverages Kolmogorov-Arnold Networks to improve interpretability and capture complex temporal dependencies in multivariate time series forecasting.

Contribution

It presents a novel transformer architecture integrating Kolmogorov-Arnold Networks for enhanced interpretability and modeling of complex temporal patterns in time series data.

Findings

Effective modeling of long-range dependencies

Improved interpretability of temporal relationships

Enhanced performance on multivariate time series forecasting

Abstract

Capturing complex temporal patterns and relationships within multivariate data streams is a difficult task. We propose the Temporal Kolmogorov-Arnold Transformer (TKAT), a novel attention-based architecture designed to address this task using Temporal Kolmogorov-Arnold Networks (TKANs). Inspired by the Temporal Fusion Transformer (TFT), TKAT emerges as a powerful encoder-decoder model tailored to handle tasks in which the observed part of the features is more important than the a priori known part. This new architecture combined the theoretical foundation of the Kolmogorov-Arnold representation with the power of transformers. TKAT aims to simplify the complex dependencies inherent in time series, making them more "interpretable". The use of transformer architecture in this framework allows us to capture long-range dependencies through self-attention mechanisms.