Copyright © 2026 Authors retain the copyright of this article. This article is an open access article distributed under the Creative Commons Attribution License which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

@article{198338,
        author = {Saikrishna L and Sreman Narayana S and P Cheenu Sriyan and Dr. R.Subhashini},
        title = {Predicting Solar Energy Yield Using Machine Learning Algorithms: A Comprehensive Approach},
        journal = {International Journal of Innovative Research in Technology},
        year = {2026},
        volume = {12},
        number = {11},
        pages = {8632-8639},
        issn = {2349-6002},
        url = {https://ijirt.org/article?manuscript=198338},
        abstract = {The world is moving towards renewable sources of energy and this has increased the imperative of precise solar energy yield predictions. The proposed research introduces a new hybrid machine learning system that takes the XGBoost feature learning and Long Short-Term Memory (LSTM) networks with attention mechanisms to forecast solar photovoltaic (PV) power output. In contrast to traditional methods that only use single models, our methodology takes advantage of the gradient boosting features of XGBoost to derive important feature importance trends that are inputted to an attention-enriched LSTM to model a temporal sequence. It was found that the proposed system was strictly tested with real meteorological data and with artificially created data with the cyclic encoding of time. Key performance indicators show better performance with an R2 of 0.96 and RMSE of 12.4 kW, significantly exceeding individual XGBoost (R 2=0.89), LSTM (R 2=0.91) and baseline ANN models found in the literature. The analysis of the feature importance indicates that solar irradiance is the most important predictor (42% importance), then temperature, and cloud cover. This mixed method solves the shortcomings of the traditional statistical models and individual ML algorithms by satisfying the spatial feature associations and temporal associations. The framework provides a practical utility when grid operators, energy traders and PV plant managers are interested in maximizing the contribution of energy dispatch, curtailment minimization, Predicting Solar Energy Yield Using Machine Learning Algorithms: A Hybrid XGBoostLSTM Approach Abstract and optimizing renewable integration. The strength and extensiveness of our solution to be applied in real-world solar forecasting are verified through the implementation details, mathematical formulations, and thorough assessment of the solution against the state of the art methods.},
        keywords = {Solar yield prediction, Long Short-Term Memory (LSTM), Hybrid Machine Learning, XGBoost, DeepSurv},
        month = {April},
        }