Sudden cardiac death (SCD) is a significant cardiovascular issue that affects approximately 3 million individuals globally each year, often occurring without any prior noticeable symptoms. The precise causes of SCD remain unclear, although ventricular fibrillation is thought to play a crucial role in its pathophysiology. Since symptoms usually appear only an hour before the event, timely prediction is essential for effective cardiac resuscitation. This study aims to predict SCD using time-frequency analysis of ECG signals. Two online datasets were utilized: the Sudden Cardiac Death Holter dataset and the MIT-BIH Normal Sinus Rhythm dataset. The proposed method involves segmenting the 60-minute interval prior to ventricular fibrillation into one-minute segments, which are then decomposed into time-frequency sub-bands using empirical mode decomposition (EMD). Nonlinear features are extracted from these decomposed signals, followed by classification using support vector machines (SVM) and K-nearest neighbors (KNN). To enhance classification accuracy, two statistical feature selection techniques, T-test and ANOVA, were employed. Results indicate that using the ANOVA feature selection method with SVM and KNN algorithms achieves high accuracy in predicting SCD. Specifically, the average accuracy rates for the 60 minutes preceding SCD were 93.51% for ANOVA-SVM and 93% for ANOVA-KNN. With T-test feature selection, the average accuracy rates were 93.29% for SVM and 93.41% for KNN. These findings demonstrate the promising performance of the proposed approach in predicting SCD.