Prediction Analysis with Professional Basket Ball Teams Using Historical Data

Abstract

I.INTRODUCTION Prediction analysis is a field that entertain in extracting information[5] from a given dataset and use it to predict future results and further performance patterns. This technique is usually applied on past or present data whereas the predictionresult is in future. For Instance, recognizing traces after anydamage has been committed,or credit card theft, after it occurred. However, main key factor is, that closeness of results are directly proportional to principle of analysis and accuracy of supposal. Evaluating at a more granular level can be determined as Predictive analytics.Tsai [7] proved that in the field, a team's performance is directly proportional to effectiveness of three major positions (i.eGuard, Forward and Centre).Min et al.[13] provided a mechanism for forecasting Football Match resultsby making use of in-game time series, Bayesian Inference and a rule-based reaasoning. His mechanism provided high precision results.Wu[8] performed a study on BasketBall Matches held in High Schools and came to an observation that ranking of teams will always go down with the increase in personal fouls.The Ranking goes up with the acceleration in total rebounds and blocks between playing teams. Performance also goes up with increase in two-point shot hit rate.S trumbelj and Vracar[12] prepared a model for predicting basketball match result.Their results of model came out to be far accurate as compared with commonly used traditional methods.Kvan and Sokol[16] devised a combinational model using Markov chainfor predicting plus using Regressionfor forecasting basketball matches held in colleges. It proved instumentaland gave more precise outcomes as compared to other Mathematical models,existing prediction logics, tournament seeding etc.Mai [9] determined that players must reduce the personal as well as turnover mistakes known as fouls in game to win it. Kvan and Sokol[10] devised a combinational model using Markov chainfor predicting plus using Regressionfor forecasting basketball matches held in colleges. It proved instumental and gave more precise outcomes as compared to other Mathematical models,existing prediction logics, tournament seeding etc.Mai [14] determined that players must reduce the personal as well as turnover mistakes known as fouls in game to win it. Lee and Worthington[20] made use of data analysis to calculateaccuracy of National Basketball Association guards, their results were greatly remarkable.Chen and Sun[15] prepared a multiple-factor decision-making structure by combiningevaluation laboratory having analytical network process methodsas well as decision making attempts to decipher significant criteriaand importance of individual factor and the surrounding environmental impact on the matchLvankovic et al. [19] detertmined that with the increase in number of features, the classification accuracy is also increased.Liu [11] analysed that a higher level of Offensive capabilityproduces a higher Rank. He also claimed that Predition of Ranking can be easily calculated withthe avera...