Abstract: This research proposes a hybrid multi-criteria decision-making (MCDM) framework using Fuzzy TOPSIS and Fuzzy ELECTRE to evaluate renewable energy alternatives. XGBoost was found to be the most predictive model, with explainability provided by SHAP and DALEX.
Abstract: This study investigates chaotic transformations that affect the performance and interpretability of AI models in complex systems. Results showed that Rossler chaotic system and CatBoost algorithm yielded 99% accuracy.
Abstract: This study introduces a reinforcement learning framework for exoplanet detection, utilizing the Advantage Actor-Critic (A2C) algorithm to identify planetary transits in light curve data.
Abstract: This research proposes a hybrid framework combining classical ANNs and Quantum Neural Networks (QNN) optimized by Genetic Algorithms for the inverse kinematics of industrial robots, achieving a 17.2% reduction in Mean Squared Error.
Abstract: Examined the potential of artificial neural networks (ANN) for calculating inverse kinematics of SCARA robot arms, achieving an R2 score of 0.93.
Abstract: This study used genetic algorithms for hyperparameter optimization of RandomForest models, with trajectory parameters explained via SHAP.
Abstract: Implemented a PID-based torque control system for a 2-DOF polar robot arm, ensuring high precision and faster reference tracking.
Abstract: Discusses the effective use of fuzzy logic control to adapt a robotic arm to variables like distance and angle during object grasping.