Heart disease remains one of the leading causes of morbidity and mortality worldwide, highlighting the need for accurate and efficient diagnostic tools. In this study, we focus on the optimization of classification models for predicting coronary heart disease through the integration of feature selection and dimensionality reduction techniques. The objective is to enhance model performance while reducing the complexity and redundancy of input features. We evaluated the performance of four machine learning algorithms: Logistic Regression, Decision Tree, Random Forest, and AdaBoost. The models were tested with and without feature selection techniques, including SelectKBest (ANOVA F-value and mutual information) and Recursive Feature Elimination (RFE), as well as dimensionality reduction through Principal Component Analysis (PCA). The experiments were conducted using a heart disease dataset, and model performance was assessed based on accuracy, precision, recall, F1-score, ROC curve, and AUC metrics. Our results demonstrate that integrating feature selection techniques significantly improves model accuracy and robustness compared to models without feature optimization. Notably, PCA proved highly effective when combined with Random Forest and Logistic Regression, achieving superior classification performance. These findings highlight the importance of selecting relevant features to improve predictive accuracy in heart disease classification tasks, paving the way for more precise and reliable diagnostic tools in clinical practice.