For Big Data, the term "machine learning" must be reinvented. In this work, we give a survey on machine learning for large data processing in ongoing reviews, as well as a recommended approach for evaluating huge data. This examination provides a point of view on the area, identifies research gaps and opportunities, and provides a reliable foundation and encouragement for further research in the field of machine learning with Big Data through this process. We looked at various data kinds, learning methodologies, important difficulties in big data management, and the usage of machine learning algorithms in big data. To increase prediction accuracy, we suggested a decision tree merging process and an ODTA (Optimistic Decision Tree Algorithm) approach in this paper.