Artificial Neural Network-Salp-Swarm Algorithm for Stock Price Prediction

Abstract

Predicting stock prices is a challenging task due to the numerous factors that impact them. The dataset used for analyzing stock prices often displays complex patterns and high volatility, making the generation of accurate predictions difficult. To address these challenges, this study proposes a hybrid prediction model that combines the salp-swarm algorithm and the artificial neural network (SSA-ANN). The SSA is used to optimize the weights and biases in the ANN, resulting in more reliable and accurate predictions. Before training, the dataset is normalized using the min-max normalization technique to reduce the influence of noise. The effectiveness of the SSA-ANN model is evaluated using the Yahoo stock price dataset. The results show that the SSA-ANN model outperforms other models when applied to normalized data. Additionally, the SSA-ANN model is compared with other two hybrid models: the ANN optimized by the Whale Optimization Algorithm (WOA-ANN) and Moth-Flame Optimizer (MOA-ANN), as well as a single model, namely the Autoregressive Integrated Moving Average (ARIMA). The study’s findings indicate that the SSA-ANN model performs better in predicting the dataset based on the evaluation criteria used.