The Indian Premier League is a professional Twenty20 cricket league, contested by eight teams based out of eight different Indian cities.
A team can acquire players through any of the three ways:
• The annual player auction
• Trading players with other teams during the trading windows, and
• Signing replacements for unavailable players.
The idea of the project is to identify and predict the value of each IPL player, based on certain obvious features that significantly affect the bid value of the player in the IPL group.
This, prediction of the value will also give an idea of what the maximum value a player could be asked for as a bid, rather than bidding a considerably higher amount all on just one player and then running short while getting to choose the other players for the team.
The Cat boost Regressor model that is built gives the best accuracy of 89% and with a lower loss function of RMSE value, compared to any other model. Thus, the above model can be used to predict any IPL Player's value.
More Hyper tuning or Feature engineering can be carried over to increase the accuracy of the predicting model