Theoretically, the output of the wind turbine might be estimated based the most known power equation that depends mainly in the wind speed. There are many issues appeared in the phase of the estimating and control while applying this equation due to ignoring many weather conditions. This paper introduces a multivariate estimation for the power curve of the wind turbine considering the weather conditions such as wind speed, air density, wind turbulence, and wind share. There variables are termed features, and a lot of measurement has been occurred to collect all possible data for these features, where measurements (system data) exceed 47,000 points. this data is proceeded mainly by three steps of the data sciences; exploratory data analysis (EDA), data processing, and building the model, using Python programming language, where it gives more flexibility more than the other languages. The power curve estimation is executed using different machine learning tools such as linear regression, polynomial regression, random forest regression, gradient boost (G Boost), and extreme gradient regression (XGBoost). A comparative study is introduced considering the R-square and the root mean square error. From the results, XGBoost learning tool provides the best performance in terms of root mean square error (RMSE). The RMSE value decreases to 6.404 while using the proposed algorithm compared to (6.631, 6.6721, and 9.072) attained through the alternative G Boost, forest random, and 4th-degree polynomial respectively.