Objective**: Implement an Artificial Neural Network (ANN) for cardiac signal

classification using either PyTorch or Keras. Follow the steps outlined below to complete the assignment.

Step 1:

Dataset Loading: Load the provided cardiac signal dataset using ‘np.load’. Dataset Link:

Step 2:

Data Preparation: Combine the two files into a single dataset. Append file into one.

Step 3:

Signal Preprocessing: Apply the provided preprocessing function to the dataset:

Step 4:

Ground Truth Creation: Generate Y_data (ground truth) for classification: Y_data=[0]*200+[1]*200

Step 5:

Data Splitting: Divide the dataset into training and testing sets.

Step 6:

Model Construction: • Design and implement an ANN model with a minimum of 5 layers. • Ensure that you use the ReLU activation function. ##Step 7: Training and Testing: Evaluate the model by testing it for both 10 and 50 epochs.

Step 8:

Model Saving: Save the trained model. ##Step 9: Evaluation Metrics: Generate a confusion matrix. Calculate precision, recall, and accuracy. Signal Preprocessing: use this. def augtime(xdatat): X_data=[] for i in range(0,len(xdatat)): if len(xdatat[i])<20000: tem=np.append(xdatat[i],xdatat[i][0:(20000-len(xdatat[i]))]) while len(tem)<20000: tem=np.append(tem,xdatat[i][0:(20000-len(tem))]) else: tem=xdatat[i] X_data=np.append(X_data,tem,axis=0) X_test=np.reshape(X_data,(len(xdatat),20000,1)) return X_test DataSet Details: • The dataset consists of normal and abnormal cardiac sounds. • Sample rate: 14,000 Hz. • Butterworth filter settings: Low-cut: 200 Hz, High-cut: 20 Hz.