In recent years, speech recognition has gained significant importance in Human-Computer Interaction (HCI). Accurately identifying the gender of a speaker based on their speech has valuable implications for applications like speech-based user authentication and personalized services. This research study presents a novel decision tree algorithm for gender identification using speech features. The algorithm utilizes a dataset of speech recordings from male and female speakers to extract features such as pitch, formants, and energy from each recording. These features are employed to train a decision tree model, capable of classifying new speech recordings as male or female. The algorithm's performance is evaluated using cross-validation, and it is compared with other commonly used machine learning algorithms. Results demonstrate the decision tree algorithm's superior accuracy rate of 93%, which outperforms support vector machine and k-nearest neighbor algorithms. Moreover, this study investigates the significance of different speech features in gender classification, highlighting pitch and formants as the most informative features. Additionally, the further insights into the decision-making process of the algorithm are obtained through the analysis of decision tree model. Overall, the proposed decision tree algorithm offers a reliable and effective approach for gender identification based on speech features. It holds potential applications in various domains, including speech-based authentication, personalized services and gender recognition in the field of human-robot interaction