Lists
Lists
Kumar, Ashish, Hassan, Bilal and Wasiq, Muhammad Farooq (2024) CricPredict: resource-aware prediction of T20 cricket match. In: 2024 International Conference on Electrical Engineering and Computer Science (ICECOS 2024), 25-26 September 2024, Palembang, Indonesia. (In Press)
One of the key problems in cricket is the increasing number of abandoned matches due to unusual circumstances. There is a total of three different formats in cricket e.g., Test, ODI and T20 international. Usually, the Duckworth–Lewis (D/L) method is used to decide the outcome of the match in Test and ODI cricket, resulting in favour of one team like completed matches. In contrast to the traditional D/L method, we tried to incorporate players' performance indicators into our proposed architecture despite the traditional D/L method which only includes the current state of the match and determines the outcome. To accomplish this task, we tried multiple different machine learning techniques e.g., Support Vector Machine (SVM), K-Nearest Neighbors (KNN), Naïve-Bayes, Linear Regression and Polynomial Regression and a deep learning model to predict the outcome of the match. To train and validate our developed architecture, we crawled data from the Indian Premier League (IPL) for the completed matches. Our proposed architecture takes complete matches as input and for the second batter, it predicts outcome at intermediate stages of matches. Later, the performance of our proposed architecture is computed using different performance indicators e.g., accuracy, Mean squared error etc. In our opinion, our proposed resource-aware prediction architecture is a unique contribution of its kind in the field.
Available under License Creative Commons Attribution 4.0.
Download (770kB) | Preview
 |
View Item |