Traditional disease diagnosis methods often involve time-consuming and costly procedures. This study presents a ML technique for symptom-driven disease prediction, leveraging statistical patterns in user-reported symptoms. The objective is to classify health conditions into disease categories with high accuracy using widely available symptom data. A comparative evaluation of six machine learning algorithms—Support Vector Classifier (SVC), K-Nearest Neighbors (KNN), Decision Tree, Gaussian Naïve Bayes, Random Forest, and Logistic Regression—was conducted. The dataset, comprising 67,200 symptom-disease pairs, was pre-processed using one-hot encoding to represent symptoms numerically. Experimental results indicate that SVC achieved the highest accuracy of 98.54%, outperforming other models in precision, recall, and F1-score. This research demonstrates the feasibility of AI-driven preliminary diagnosis and suggests future integration with telemedicine platforms for cost-effective and accessible healthcare solutions.