Deep Learning for ECG Signal Analysis

This demo shows how can we build a model using transfer learning techniques to classify signal data. Here we used Wavelet Transform to generate a time-frequency representation of a signal. The Wavelet Transform greatly enhances the spectral information present as a function of time in the signal and we can save this information as a Image.

Step 1:

Explore the Signals.

Step 2:

Generate Wavelet Time-Frequency Representations.

Step 3:

Train a Pre-Trained Network (AlexNet).

Step 4:

Evaluate the Trained Model.