Poll is CLOSED
Poll
I have coded up a Deep Learning/Neural Network program.
Yes in MATLAB
18%
Yes but only in some other language
22%
Yes in MATLAB and other language(s)
7%
Not yet but would like to learn how
34%
No & don't foresee any need to yet
19%
4243 votes
7 Comments
If you answered "Yes, but only in some other language", or "Yes in MATLAB and other language(s)", you might be interested to know that you can use MATLAB with TensorFlow and PyTorch for deep learning:
- You can import TensorFlow and PyTorch models into MATLAB. You can do that to combine preprocessing and or postprocessing algorithms in MATLAB with AI algorithm that was developed in TensorFlow or PyTorch. Or you can generate library-free C code from the imported model for deploying it to a microcontroller, ECU, or a PLC. You can also generate HDL to target FPGA or CUDA code to target a GPU. You can simulate imported models in Simulink to test how your AI model interacts with control and signal processing algorithms modeled in Simulink.
- You can also co-execute TensorFlow and PyTorch models with MATLAB algorithms by calling TensorFlow and PyTorch models from MATLAB directly. You can even use MATLAB to optimize hyperparameters of a deep learning model you are training in TensorFlow.
Watch a relatively short video that explains all this in more detail and ask if anything is unclear.
Adding some action items at the polling exit.
If you selected "Not yet but would like to learn how" or "No & don't foresee any need to yet", why not take a look at: https://www.mathworks.com/solutions/deep-learning/tutorials-examples.html?
And take a casual walk around the gallery: https://www.mathworks.com/help/deeplearning/examples.html
Enjoy!
The current version and future versions are increasingly facilitating some customised operations in terms of deep learning, which I did a lot of practice with a few years ago.
I'm very curious and excited to learn MATLAB (•_•)(•_•)
Just to see how it develops: about 40% of early voters voted for the 'I want to learn to' option. 'in Matlab' was 25%, with an even spread for the remainder.