TensorFlow for Hackers (Part IV) - Neural Network from Scratch

Let's build Neural Network classifier using only Python and NumPy. We will implement the Backpropagation algorithm and use it to train our model. Finally, our newly created classifier will be used to recognize digits from the MNIST dataset. »
Author's profile picture Venelin Valkov on data-science

TensorFlow for Hackers - Part III

Let's build a Convolutional Neural Network that can distinguish cats from dogs using only raw pixel values. We will use a little helpful library called TFLearn. Is it going to be good enough to correctly classify your cat or a dog? »
Author's profile picture Venelin Valkov on data-science

AppGym - Android Apps as Reinforcement Learning Environment

Is it possible to use mobile apps as a training ground for your RL agents? You might need a tad bit more than sugar, spice and everything nice to do it. »
Author's profile picture Venelin Valkov on data-science

How to get Code Coverage during Manual Testing for Android App

Are you stuck trying to do some manual testing and get the code coverage from it? Here is a simple (but somewhat barebone) solution to that problem. »
Author's profile picture Venelin Valkov on dev

TensorFlow for Hackers - Part II

In this one, you will build a Neural Network which tries to decide whether or not a student has drinking problem. Will you get good accuracy from the model? »
Author's profile picture Venelin Valkov on data-science