Integration of Wavelet Transform Convolution and Channel Attention with LSTM for Stock Price Prediction based Portfolio Allocation

TL;DR

This paper introduces a novel stock prediction model combining wavelet transform convolution, channel attention, and LSTM to improve portfolio allocation, demonstrating robustness in volatile market conditions.

Contribution

The study presents an integrated approach that enhances stock price prediction accuracy by combining noise reduction, feature reconstruction, and sequence modeling.

Findings

Achieves higher return and Sharpe ratio compared to baseline models.

Demonstrates robustness during post-pandemic market downturns.

Reduces noise in stock time series effectively.

Abstract

Portfolio allocation via stock price prediction is inherently difficult due to the notoriously low signal-to-noise ratio of stock time series. This paper proposes a method by integrating wavelet transform convolution and channel attention with LSTM to implement stock price prediction based portfolio allocation. Stock time series data first are processed by wavelet transform convolution to reduce the noise. Processed features are then reconstructed by channel attention. LSTM is utilized to predict the stock price using the final processed features. We construct a portfolio consists of four stocks with trading signals predicted by model. Experiments are conducted by evaluating the return, Sharpe ratio and max drawdown performance. The results indicate that our method achieves robust performance even during period of post-pandemic downward market.