The neurological problem in the brain causes epileptic seizures, which can have an impact on a patient's health. Electroencephalogram (EEG) signal data is provided, and machine learning (ML) techniques are employed to predict epileptic seizures from the dataset. The major concern is the artifacts and noisy removal. Using a Discrete Wavelet Transform (DWT), the preprocessing method for classifying epileptic seizures is made more computationally fast. The feature extraction cutoff is used and the Daubechies wavelet is used as a scaling function to identify the noisy data from the EEG data set. The proposed method uses Principal Component Analysis with the QR Algorithm to reduce the dimensionality of the dataset, preserving the most significant components of Logistic Regression, as well as the Convolutional Neural Network (CNN) classifier, was compared with the classification relations of EEG signal to identify epileptic seizures. By applying DWT in connection with Convolutional Neural Network obtained novel and reliable classification with accuracy. Detection of epilepsy through the proposed ML algorithm is shown in the experimental results through EEG signals that merit this study with notable parameters like specificity, sensitivity, and accuracy.