cardiovascular diseases (CVD) are found to be rampant in the populace leading to fatal death. Worldwide, the majority of people are suffering from CVD. However, the performance of current deep learning models is heavily reliant on the availability of large, well-annotated training datasets, which are frequently difficult to obtain in the medical domain due to privacy concerns and labeling costs. Inadequate or imbalanced data might cause biased predictions, lowering the model's reliability in detecting fewer common arrhythmias and possibly leading to diagnostic errors in essential cases. In this study a novel Machine learning-based CAD-SULOR model is proposed for CVD using support vector machine (SVM) and logistic regression (LR). The input signal is pre-processed using Least mean square (LMS) algorithm to reduce the noise and enhance the signal. Discrete Wavelet Transformation (DWT) is used to extract the features from the ECG signal. The SVM, and LR are utilized to improve the accuracy and classify the CVD, such as normal and abnormal. The performance of the CAD-SULOR approaches was assessed using the metrics such as F1 score, specificity, recall, accuracy, and precision. The CAD-SULOR approach achieves a high accuracy of 99.26% for cloth retrieval. The CAD-SULOR the accuracy 8.36%, 2.84% and 0.39% better than DDR-Net [13], MobileNet [14], and LSTM [16] respectively.