How can I join a machine learning model and an LSTM model into a single neural net?

Hi,

I'm training an LSTM network on a series of music chords to classify the piece of music; however, the results are not really good at all. I have previously used the machine learning tool box to analyse the piece of music soely based on the ratios of different chords, and I have achieved significantly better results with the conventional machine learning models than I have with the LSTM. My thesis advisor told me to maybe consider combining the two techniques, ie. to use the output of the machine learning model, in this case an SVM, and feed it into a neural networ. However, I don't know how to do that in the Deep Learning Application.

Answer 1

[Nagasai Bharat]

Hi,

For getting the started with Deep learning using MATLAB you can use this documentation of deep learning toolbox

Get Started with Deep Learning Toolbox (mathworks.com)

Deep Learning with Time Series, Sequences, and Text - MATLAB & Simulink (mathworks.com)

For using both neural networks and SVMs you could take the average of both prediction metrics to have a final prediction.