Abstract—The widespread adoption of Wireless Sensor Networks (WSNs) is driven by their characteristics and performance, fueling rapid growth across various sectors. However, these networks are highly susceptible to numerous security threats, with Denial-of-Service (DoS) attacks being the most common. To ensure the security and reliability of WSN services, it is crucial to implement an intrusion detection system capable of recognizing a wide range of security threats. The Orange toolbox, used in conjunction with the 10-fold cross-validation method, enhances the detection and accurate classification of different Denial of Service (DoS) attacks. A specialized dataset for WSN-DS, obtained from Kaggle, was used to train several Machine Learning (ML) models to detect and classify four types of Denial of Service (DoS) attacks, Blackhole, Grayhole, Flooding, and Scheduling attacks, as well as normal behavior. The evaluated classifiers, including Adaboost, CN2 Rule Inducer, and Random Forest (RF), achieved an accuracy rate exceeding 99.33%. The Adaboost classifier outperformed the others, achieving a perfect accuracy rating of 100%. Additionally, performance metrics such as accuracy, Fmeasure, precision, and sensitivity indicated that the Adaboost classifier consistently surpassed the other models, attaining an accuracy rating of 100%, while CN2 Rule Inducer and RF classifiers recorded accuracy rates of 99%. Further performance measures, including the confusion matrix, were utilized to assess these various ML classifiers and determine their effectiveness in accurately detecting harmful behavior. Subsequently, the three classifiers maintained a high accuracy rate.