Stock Price Forecasting Using Single Multiplicative Neuron Model Artificial Neural Network (SMNM-ANN): A Case Of Iron-Steel Sector in Türkiye

This study employs the Single Multiplicative Neuron Model Artificial Neural Network (SMNM-ANN) Automatic Forecasting Method for Stock Price Prediction in Turkey's steel industry. Accurate financial forecasts are critical for businesses and investors, as the primary goal is to maximize profits through predicting future stock prices. A variety of methods are used in stock price predictions, with traditional techniques being complemented by artificial neural networks and machine learning methods in recent times. In this context, the SMNM-ANN method provides a flexible, nonlinear approach for predicting stock prices. The study utilizes 1101 stock closing price data between January 28, 2015, to May 30, 2025. To facilitate comparisons, machine learning techniques such as Boosted Tree, Decision Tree, and SVM are also employed. The application calculates forecast values for the next 10 data points, revealing that the SMNM-ANN-AFM method exhibits lower error rates. The obtained RMSE value is 0.6983, while the MAPE value is 2.5138. This research highlights the potential of SMNM-ANN to enhance decision-making processes and resource allocation in the steel sector, thereby contributing significantly to the industry.