Stock price forecast with deep learning

TL;DR

This paper compares different neural network architectures and optimization techniques for predicting the next-day S&P 500 index value, finding that a single-layer recurrent neural network with RMSprop performs best.

Contribution

It provides a systematic comparison of neural network architectures and optimizers for stock price prediction, highlighting the effectiveness of a simple recurrent network with RMSprop.

Findings

Recurrent neural networks outperform other architectures.

RMSprop optimizer yields the lowest prediction error.

Achieved validation MAE of 0.0150 and test MAE of 0.0148.

Abstract

In this paper, we compare various approaches to stock price prediction using neural networks. We analyze the performance fully connected, convolutional, and recurrent architectures in predicting the next day value of S&P 500 index based on its previous values. We further expand our analysis by including three different optimization techniques: Stochastic Gradient Descent, Root Mean Square Propagation, and Adaptive Moment Estimation. The numerical experiments reveal that a single layer recurrent neural network with RMSprop optimizer produces optimal results with validation and test Mean Absolute Error of 0.0150 and 0.0148 respectively.