I am beginner in Nengo , I am looking for simple classification task in Nengo . Can anyone provide simple classification with 2-3 categories , suppose if classification task is text related then should i go for same approach as i used to do in deep learning ?
In deep learning i was converting text data to int values using dict , and feeding those mapped values to neural network , should i try same with nengo ?
Thank you !
Hi @Monk. Nengo is a bit more awkward to use for classification than other artificial neural network programs like TensorFlow because we usually use spiking neurons that have dynamics across timesteps, rather than each timestep being a separate classification.
It is still possible though. If your background is in machine learning, then you will likely find Nengo DL more familiar. In particular, see this MNIST classification example, which uses spiking neurons to classify MNIST digits.
Hello @tbekolay, I cannot find an example that learns MNIST(or any other classification) based on unsupervised learning. I find it a bit difficult to write the training, evaluation procedure. Are you aware of any such example?
Hi @nikhilgarg, no to the best of my knowledge there are no Nengo examples that use fully unsupervised learning to do classification. Are you thinking about biological unsupervised learning rules (e.g. STDP) or the types of unsupervised learning used in machine learning?
I am thinking of biological algorithms like STDP and BCM. Similar to diehl and cook 2015 paper.