A review of Machine Learning based Recommendation approaches for cricket

Cricket is one of the most watched sports in the world. Analyzing the factors, activities that affect the aftermath of the matches is of interest to many cricket lovers. Machine learning (ML) has demonstrated favorable outcomes in various fields like medical diagnosis, image processing, prediction, classification, regression etc. and is evidenced to be accurate. The ingenious frameworks built on ML have scope to learn from past experiences. The cricket pitch is the prime segment along with home game advantage, coin toss, innings, day/night match, physical fitness etc. A compendious examination of the cricket pitch, if done with imminent objectives, will be useful to precisely speculate outcomes match. This paper focuses on the existing research for analyzing the cricket pitch, predicting the performance of players and outcomes of the matches. A comparison among machine learning techniques has also been done and Naïve Bayes gave the best results due to its capability of producing accurate results with small samples while in other researches random forest turned out to be the precise classifier as it gave more accuracy than any other approach used.