Apollo-Forecast: Overcoming Aliasing and Inference Speed Challenges in Language Models for Time Series Forecasting

TL;DR

Apollo-Forecast introduces innovative techniques to improve time series forecasting with language models by reducing aliasing distortion and significantly speeding up inference, especially for large models.

Contribution

The paper presents Apollo-Forecast, a framework with novel AAQM and RD modules that enhance encoding fidelity and inference speed in language model-based time series forecasting.

Findings

Outperforms state-of-the-art methods by 35.41% in WQL and 18.99% in MASE metrics.

Achieves 1.9X-2.7X faster inference speed.

Effectively mitigates aliasing and accelerates long-term predictions.

Abstract

Encoding time series into tokens and using language models for processing has been shown to substantially augment the models' ability to generalize to unseen tasks. However, existing language models for time series forecasting encounter several obstacles, including aliasing distortion and prolonged inference times, primarily due to the limitations of quantization processes and the computational demands of large models. This paper introduces Apollo-Forecast, a novel framework that tackles these challenges with two key innovations: the Anti-Aliasing Quantization Module (AAQM) and the Race Decoding (RD) technique. AAQM adeptly encodes sequences into tokens while mitigating high-frequency noise in the original signals, thus enhancing both signal fidelity and overall quantization efficiency. RD employs a draft model to enable parallel processing and results integration, which markedly accelerates the inference speed for long-term predictions, particularly in large-scale models. Extensive experiments on various real-world datasets show that Apollo-Forecast outperforms state-of-the-art methods by 35.41\% and 18.99\% in WQL and MASE metrics, respectively, in zero-shot scenarios. Furthermore, our method achieves a 1.9X-2.7X acceleration in inference speed over baseline methods.