Thresholded ConvNet Ensembles: Neural Networks for Technical Forecasting

TL;DR

This paper demonstrates that an ensemble of thresholded convolutional neural networks trained on financial time series data can outperform traditional technical analysis methods in stock market forecasting, offering improved accuracy and interpretability.

Contribution

It introduces a novel ensemble of shallow, thresholded CNNs optimized over multiple resolutions for financial forecasting, surpassing traditional technical analysis performance.

Findings

Ensemble CNNs outperform traditional technical methods.

Learned filters provide visual interpretability.

Multi-resolution approach enhances predictive accuracy.

Abstract

Much of modern practice in financial forecasting relies on technicals, an umbrella term for several heuristics applying visual pattern recognition to price charts. Despite its ubiquity in financial media, the reliability of its signals remains a contentious and highly subjective form of 'domain knowledge'. We investigate the predictive value of patterns in financial time series, applying machine learning and signal processing techniques to 22 years of US equity data. By reframing technical analysis as a poorly specified, arbitrarily preset feature-extractive layer in a deep neural network, we show that better convolutional filters can be learned directly from the data, and provide visual representations of the features being identified. We find that an ensemble of shallow, thresholded CNNs optimised over different resolutions achieves state-of-the-art performance on this domain, outperforming technical methods while retaining some of their interpretability.