Human activity recognition (HAR) has garnered considerable interest in recent years owing to its various applications in domains like healthcare, sports, and smart environments. Analyzing and classifying human activities from sensor data, such as accelerometers and gyroscopes, provide valuable insights into understanding behavior patterns and enhancing user experiences. In this research, we focus on the critical task of optimizing hyperparameters for the human activity recognition dataset using machine learning (ML) with the random forest classifier as the base model. UCI machine learning repository’s human activity recognition dataset leverages wearable sensor technologies, which capture rich temporal information about various activities. The dataset's complexity and diversity present a challenging scenario, necessitating sophisticated machine learning models with well-tuned hyperparameters to achieve optimal performance. Hyperparameters play a vital role in enhancing the performance of ML and DL models, influencing factors such as model capacity, regularization, and convergence. The exploration of hyperparameters is a critical step in maximizing model accuracy and generalization. Traditional methods involve manual tuning through trial and error, a time-consuming and resource-intensive process. However, the rise of automated hyperparameter optimization techniques provides an efficient alternative. This research employs the random forest classifier, a model that is versatile and extensively utilized, as the foundation for evaluating different hyperparameter optimization algorithms.