Sleeping Disorder in Bangladeshi Female Cricketers: A Machine Learning Perspective

Abstract

The Sleep quality really matters when it comes to how athletes perform and recover. Sports science pays a lot of attention to this. The study here looks at sleep in Bangladeshi female cricketers. It uses machine learning to predict sleep results and figure out what affects them. Data came from 146 athletes. Features included age, years of experience, how intense their practice was, BMI, and indicators of sleep behavior. The Athlete Sleep Behavior Questionnaire gave a continuous sleep score. That was the main target for prediction. Preprocessing happened first. Things like mean imputation and feature engineering. Then exploratory analysis showed some clear patterns. Experience, practice level, and age all strongly shaped sleep quality. Athletes who had more experience and ramped up their practice tended to sleep better overall. Four machine learning models got tested. Linear Regression, Random Forest, XGBoost, and Artificial Neural Networks, or ANN. Random Forest did okay with an RMSE of 4.35 on the test data. It pointed to training intensity and experience as the big predictors. But the ANN model came out on top. Its RMSE was just 0.47. Age turned out to be the top factor there. Experience and practice level followed close behind. Sensitivity analysis backed this up. It confirmed age as the main driver. This lines up with what evidence suggests. Structured training and getting older seem to help with better recovery and sleep. The research highlights why tracking sleep patterns in athletes is so key. It also shows how machine learning can dig into what determines sleep quality. Findings point to ways to improve things. Optimizing training schedules and pushing sleep hygiene might boost performance and health. Future studies could bring in wearable tech and bigger datasets. That would allow for more accurate, real-time predictions of sleep.