Comparative Analysis of Various Machine Learning Techniques based on Performance Metric used in Business Forecasting

A Comparative Study of Machine Learning Techniques for Business Forecasting

Abstract

Once in the past authority judgment is the primary decision for business forecasting. It is a dull and slanted to error process. Continuously the interest is moved from authority conclusion to unadulterated genuine based forecasting models. Any business association uses bits of knowledge to characterize fitting game plans and plans. Bits of knowledge are in like manner supportive in forecasting purpose. As quantifiable speculations end up being more erratic, it is difficult to do estimations. In like manner authentic methodologies can't be insightful, nor can the rationale of framework be retained. For whatever period of time that there are no different features in the data set and relationship among the factors are immediate it is clear and alluring to use true based forecasting model. In any case, if the hidden relationship among the factors is unidentified or not immediate, machine learning based procedures are favored better. So the eagerness of making forecasting model has been moved from quantifiable field to computer field. Machine learning gives us computerized techniques for data examination. In this paper we discussed about machine learning techniques used in business forecasting and studied their performance in predicting Foreign Exchange rate.