Design of a Predictive Model Based on SMA-Multimachine Learning Algorithms

This paper describes the design of a predictive model using the Slime Mold Optimization Algorithm (SMA) in conjunction with multiple machine learning algorithms. The model is a two-stage system where the first stage estimates the probability of success and the second stage determines the exact number of success conditions. In the first stage, we compare Random Forest (RF), Support Vector Machine (SVM) and Deep-Onet models for parameter optimization using SMA. The Deep-Onet model that achieved the highest <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\mathbf{F 1}$</tex> score of <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\mathbf{0. 9 5}$</tex> was selected. In the second stage, the RF model is again used, which achieves the highest fit coefficient <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\mathbf{R}^{\mathbf{2}}$</tex> of <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\mathbf{0. 9 4}$</tex>. The resulting twostage SMA-Deep-Onet-RF model is then applied to the forecasting results, demonstrating its validity through a case study similar to the potential 2028 forecast.