This paper presents a study on the analysis and detection of electrical faults in photovoltaic (PV) arrays using machine learning classifiers. The research focuses on the challenges of detecting and classifying faults in PV systems, particularly those with similar I-V curves, which are difficult to distinguish using traditional methods. The study explores the use of Support Vector Machines (SVMs) and Extreme Gradient Boosting (XGBoost) for fault detection in small PV arrays. Additionally, optimization algorithms such as the Bees Algorithm (BA) and Particle Swarm Optimization (PSO) are employed to enhance the performance of these classifiers by tuning their hyperparameters for improved fault classification accuracy. The study constructs a small-scale PV array to simulate real-world conditions and generate a high-quality dataset for training machine learning models. The array consists of five parallel-connected strings, each with twenty modules in series. The I-V curves of the array are tested under standard test conditions (STCs) to identify various fault scenarios, including line-to-line (LL) faults and open-circuit (OC) faults. The results show that the Bees Algorithm outperforms the Particle Swarm Optimization algorithm in optimizing SVM and XGBoost classifiers for fault detection in PV arrays. The study also evaluates the performance of SVM and XGBoost classifiers using metrics such as classification accuracy, sensitivity, and specificity. The results demonstrate that the Bees Algorithm significantly improves the accuracy of these classifiers in detecting complex faults in PV arrays. The study proposes a novel fault detection setup using optimized SVM and XGBoost classifiers, which can effectively distinguish between LL and OC faults with minimal sample requirements. The findings highlight the potential of the Bees Algorithm in enhancing the accuracy of fault detection in photovoltaic systems.This paper presents a study on the analysis and detection of electrical faults in photovoltaic (PV) arrays using machine learning classifiers. The research focuses on the challenges of detecting and classifying faults in PV systems, particularly those with similar I-V curves, which are difficult to distinguish using traditional methods. The study explores the use of Support Vector Machines (SVMs) and Extreme Gradient Boosting (XGBoost) for fault detection in small PV arrays. Additionally, optimization algorithms such as the Bees Algorithm (BA) and Particle Swarm Optimization (PSO) are employed to enhance the performance of these classifiers by tuning their hyperparameters for improved fault classification accuracy. The study constructs a small-scale PV array to simulate real-world conditions and generate a high-quality dataset for training machine learning models. The array consists of five parallel-connected strings, each with twenty modules in series. The I-V curves of the array are tested under standard test conditions (STCs) to identify various fault scenarios, including line-to-line (LL) faults and open-circuit (OC) faults. The results show that the Bees Algorithm outperforms the Particle Swarm Optimization algorithm in optimizing SVM and XGBoost classifiers for fault detection in PV arrays. The study also evaluates the performance of SVM and XGBoost classifiers using metrics such as classification accuracy, sensitivity, and specificity. The results demonstrate that the Bees Algorithm significantly improves the accuracy of these classifiers in detecting complex faults in PV arrays. The study proposes a novel fault detection setup using optimized SVM and XGBoost classifiers, which can effectively distinguish between LL and OC faults with minimal sample requirements. The findings highlight the potential of the Bees Algorithm in enhancing the accuracy of fault detection in photovoltaic systems.