This paper introduces an advanced localization algorithm for Wireless Sensor Networks (WSNs) called Enhanced Distance Vector Hop with Machine Learning (EDV-ML), addressing key limitations of the traditional Distance Vector Hop (DV-Hop) algorithm. The proposed EDV-ML algorithm employs a supervised learning model to correct average hop distances based on network parameters, significantly improving accuracy. Additionally, it integrates reinforcement learning techniques to optimize node coordinates, ensuring reduced localization error. Comprehensive MATLAB simulations highlight the superior performance of EDV-ML compared to traditional DV-Hop and its variants, demonstrating substantial enhancements in positioning accuracy. The proposed approach achieves these improvements without requiring additional hardware or extensive modifications, offering a practical and efficient solution for real-world WSN deployments. This work makes a meaningful contribution to localization algorithms by combining machine learning and optimization techniques to enhance accuracy and robustness.