Time Series Forecasting Using Hybrid CNN - RNN

A hybrid convolutional neural network - recurrent neural network (RNN) for time series prediction is implemented.

Version 1.0.13 (566 KB)
3.8K Downloads
(7)
27 May 2021

You are now following this Submission

This example aims to present the concept of combining a convolutional neural network (CNN) with a recurrent neural network (RNN) to predict the number of chickenpox cases based on previous months.
The CNN is an excellent net for feature extractions while a RNN have proved its ability to predict values in sequence-to-sequence series. At each time step the CNN extracts the main features of the sequence while the RNN learn to predict the next value on the next time step.
Please rate this contribution if you think that in some how it helps you. Thank you.

Others Also Downloaded

CNN-LSTM Time Series Prediction Matlab Univariate Time Serie

495 Downloads

-- / 5 (0)

Photovoltaic Power Estimation Using DBN, LSTM, CNN & GRU

118 Downloads

5.00 / 5 (1)

Python-and-MATLAB-RNN-LSTM-Model-for-Prediction-and-Forecast

257 Downloads

-- / 5 (0)

Cite As

H Sanchez (2026). Time Series Forecasting Using Hybrid CNN - RNN (https://www.mathworks.com/matlabcentral/fileexchange/91360-time-series-forecasting-using-hybrid-cnn-rnn), MATLAB Central File Exchange. Retrieved .

General Information

MATLAB Release Compatibility

  • Compatible with any release

Platform Compatibility

  • Windows
  • macOS
  • Linux
Open in new tab
Version Published Release Notes
1.0.13

A hybrid Xception - RNN network is included in this new version. Notice Xception requires high computing power for training. Using last Gen AMD processor and last gen Nvidia 3080 may take around 30 mins to train the network.
1.0.12

A feedforward network options for prediction has been included. It is shown the process of data preparation process for a univariate time series forecasting.

1.0.11

A hybrid ResNet50-RNN was included in this example.
The example includes horizon to predict the values beyond the testing data. It now predicts unobservable values beyond the limit of the testing sequence.
Better documented and explained.
1.0.2

A hybrid ResNet50-RNN was included in this example.
The example includes horizon to predict the values beyond the testing data. It now predicts unobservable values beyond the limit of the testing sequence.
Better documented and explained.
1.0.1

A hybrid ResNet50-RNN was included in this example.
The example includes horizon to predict the values beyond the testing data. It now predicts unobservable values beyond the limit of the testing sequence.
Better documented and explained.

1.0.0