Predicting the Outcome of a T20 Cricket Game Based on the Players’ Abilities to Perform Under Pressure
| Autor: | Jitendra Sai Kota, Mamatha Vayelapelli |
|---|---|
| Rok vydání: | 2020 |
| Předmět: |
Training set
biology Computer science business.industry League Machine learning computer.software_genre biology.organism_classification Outcome (game theory) Random forest Support vector machine Cricket Signal Processing Game based Artificial intelligence Electrical and Electronic Engineering business computer Test data |
| Zdroj: | IEIE Transactions on Smart Processing & Computing. 9:230-237 |
| ISSN: | 2287-5255 |
| Popis: | With the advent of the Twenty20 format in cricket, the game has become more competitive. The number of nail-biting finishes that go to the last over has also increased as a result. As the game goes to the last over, the match result is mostly dependent on the effectiveness of batsmen at the crease and the player who is bowling the last over. This gave rise to an interesting idea for predicting the game result before the start of the final over based on the capabilities of the batsmen and the bowler. We used data from the first eight Indian Premier League (IPL) seasons as the training data, and from the last two IPL seasons as the test data. We trained and tested our model using Random Forest, Support Vector Machine (SVM) and k-nearest neighbors algorithms, and obtained the best accuracy using the SVM (89.36%). |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|