Gold Price Prediction Using Long Short-Term Memory and Multi-Layer Perceptron with Gray Wolf Optimizer

TL;DR

The paper tackles volatile gold price forecasting under multifaceted macroeconomic drivers. It proposes a three-network AI pipeline—two LSTMs for daily and monthly forecasts and an MLP fusion network—with Gray Wolf Optimizer tuning to produce next-day high/low/close predictions. Evaluation uses data from 2010–2021 across macroeconomic and market indicators and includes a live-demo trading test, achieving a daily MAE around $0.21 and monthly RMSE around $26, with a three-month demo return near 173%. The work demonstrates a practical end-to-end forecasting-to-trading workflow and highlights potential extensions to other assets and signals.

Abstract

The global gold market, by its fundamentals, has long been home to many financial institutions, banks, governments, funds, and micro-investors. Due to the inherent complexity and relationship between important economic and political components, accurate forecasting of financial markets has always been challenging. Therefore, providing a model that can accurately predict the future of the markets is very important and will be of great benefit to their developers. In this paper, an artificial intelligence-based algorithm for daily and monthly gold forecasting is presented. Two Long short-term memory (LSTM) networks are responsible for daily and monthly forecasting, the results of which are integrated into a Multilayer perceptrons (MLP) network and provide the final forecast of the next day prices. The algorithm forecasts the highest, lowest, and closing prices on the daily and monthly time frame. Based on these forecasts, a trading strategy for live market trading was developed, according to which the proposed model had a return of 171% in three months. Also, the number of internal neurons in each network is optimized by the Gray Wolf optimization (GWO) algorithm based on the least RMSE error. The dataset was collected between 2010 and 2021 and includes data on macroeconomic, energy markets, stocks, and currency status of developed countries. Our proposed LSTM-MLP model predicted the daily closing price of gold with the Mean absolute error (MAE) of $ 0.21 and the next month's price with $ 22.23.

Figures (6)

• Figure 1: Prediction steps of the proposed algorithm: data collection, preprocessing, LSTM-based daily and monthly forecasting, MLP fusion, and trading strategy generation.
• Figure 2: Architecture of the proposed LSTM--MLP algorithm. Two LSTM networks handle daily and monthly forecasts, respectively. Their outputs feed into an MLP fusion network that produces final predictions for high, low, and close prices.
• Figure 3: Optimization of neuron counts for the daily LSTM subnetworks: High (left), Close (middle), Low (right).
• Figure 4: Optimization of neuron counts for the monthly LSTM subnetworks: High (left), Close (middle), Low (right).
• Figure 5: Optimization of neuron counts for the MLP fusion subnetworks: High (left), Close (middle), Low (right).