Maybe we can start of with Softmax Regression on TensorFlow to start off with? It is very simple and can be completed quickly also. It gives a solid introduction to the basic of classification and softmax function.
But instead of using the cliched MNIST dataset, we can try and look into the alphabets one maybe?
sumodm
commented
8 years ago
Maybe we can start of with Softmax Regression on TensorFlow to start off with? It is very simple and can be completed quickly also. It gives a solid introduction to the basic of classification and softmax function. But instead of using the cliched MNIST dataset, we can try and look into the alphabets one maybe?