Autism spectrum disorder (ASD) in individuals with fragile X syndrome (FXS) presents significant diagnostic challenges due to overlapping behavioral and cognitive characteristics. Early and accurate detection is essential for timely intervention and improved quality of life. A machine learning-based approach enhances the predictive accuracy of ASD in FXS-affected individuals. Random Forest detects FXS, leveraging its ability to handle complex, high-dimensional data, while Support Vector Machine (SVM) classifies ASD cases within the FXS population. The model is trained on survey datasets incorporating Autism Quotient (AQ) scores, offering a non-invasive and efficient diagnostic tool. By utilizing these advanced classification techniques, the combined use of Random Forest and SVM provides reliable predictive performance, supporting the need for machine learning-driven diagnostic frameworks in neurodevelopmental research.