This paper explores the application of machine learning (ML) techniques for forecasting temperature and solar irradiance in photovoltaic (PV) systems. The authors review various ML algorithms, including decision trees, random forests, support vector machines (SVM), and XGBoost, and compare their performance in predicting these critical parameters. The study emphasizes the importance of accurate weather forecasts for optimizing PV system performance and grid integration. Traditional meteorological models are noted for their limitations, while ML models offer improved accuracy and reliability. The paper presents a case study using historical data from a solar farm in Hassi R'mel, Algeria, to evaluate the models' effectiveness. The results show that decision trees achieve the highest accuracy in temperature prediction, while XGBoost performs best in solar irradiance forecasting. The study concludes that ML models, particularly decision trees, have significant potential in enhancing the efficiency and sustainability of solar energy production. Future research directions include further model improvements, integration of advanced feature engineering, and the use of privacy-preserving techniques to protect sensitive data.This paper explores the application of machine learning (ML) techniques for forecasting temperature and solar irradiance in photovoltaic (PV) systems. The authors review various ML algorithms, including decision trees, random forests, support vector machines (SVM), and XGBoost, and compare their performance in predicting these critical parameters. The study emphasizes the importance of accurate weather forecasts for optimizing PV system performance and grid integration. Traditional meteorological models are noted for their limitations, while ML models offer improved accuracy and reliability. The paper presents a case study using historical data from a solar farm in Hassi R'mel, Algeria, to evaluate the models' effectiveness. The results show that decision trees achieve the highest accuracy in temperature prediction, while XGBoost performs best in solar irradiance forecasting. The study concludes that ML models, particularly decision trees, have significant potential in enhancing the efficiency and sustainability of solar energy production. Future research directions include further model improvements, integration of advanced feature engineering, and the use of privacy-preserving techniques to protect sensitive data.