Python

TensorFlow 2 has finally became available this fall and as expected, it offers support for both standard CPU as well as GPU based deep learning. Since using GPU for deep learning task has

After transforming 1D time domain data series into frequency 2D maps in part 1 of this miniseries, we’ll now focus on building the actual Convolutional Neural Network binary classification model. The goal is

This example explores the possibility of using a Convolutional Neural Network(CNN) to classify time domain signal. The fundamental thesis of this work is that an arbitrarily long sampled time domain signal can be

The TensorFlow Estimators API is great for quickly building your custom models, but you might have noticed there is no obvious way to pass custom parameters from your estimator into the input_fn. That

It is easy to hit resource limits when working with large datasets. The available memory in particular becomes quickly a limiting factor when training your neural networks on swaths of data. The solution

We’ve all been there - it’s Sunday evening, you have a couple of fresh ideas for a new customer centric strategy and you want to test how it would hold up in the

When I stumble on an interesting new dataset, I often find myself excitedly prototyping a quick machine learning models to see what type of insights I could get out of the latest find.