Prediction in machine learning is a blooming area in the computer field. It is used in almost all areas in order to predict future events. Such prediction will help people to have knowledge about the future, which will enable them to make better decisions. There are several prediction methods that have been proposed by many researchers for various applications. The performance of those systems is pretty good and is helping people to make good decisions. In this paper, we have studied various prediction systems which used machine learning techniques. The performance of these techniques is good but the problem is that the accuracy of the prediction mechanism is greatly affected by various factors such as less datasets, not being implemented in real time, using limited number of parameters, etc. We have proposed a framework which can be used to increase the accuracy of the prediction process by adding several features. Using this proposed system, accuracy of prediction results can be increased. To analyze the efficiency of the proposed framework, we used a web-based interface. In this system we predicted agriculture product price. To predict the price of the product, we used linear regression, random forest.