In daily activities generate a large amount of information from many fields, including business, economic, social networking sites, IoT, and store it on different repository or storages, etc. This information becomes important assets for our departments in future decision making. In this paper a simple step by step solution proposed for big data. Also, investigate to find how much machine learning reliable for it. So, a large data set is used for experiment in this work. Sample machine learning techniques examined for data analysis. For classification and prediction performed the XGBoost model. We obtained high accuracy 1.00 % F1 score. This study motivated researchers that machine learning is the best solution for big data analysis.