We cover basic mistakes that can lead to unnecessary copying of data and memory allocation in NumPy. We further cover NumPy internals, strides, reshaping, and transpose in detail.
In this part we'll see how to speed up an implementation of the k-means clustering algorithm by 70x using NumPy. We cover how to use cProfile to find bottlenecks in the code, and how to address them using vectorization.
In Part 1 of our series on writing efficient code with NumPy we cover why loops are slow in Python, and how to replace them with vectorized code. We also dig deep into how broadcasting works, along with a few practical examples.
With countless options to design neural networks, an effective architecture search algorithm would be game-changing. Here we look at the state of the art.
Here we see how to speed up NumPy array processing using Cython. By explicitly declaring the "ndarray" data type, your array processing can be 1250x faster.
Python might be one of today's most popular languages, but it's definitely not the most efficient. See how Cython can easily boost your Python scripts.
This is the second tutorial in the series which discusses extending the implementation for allowing the GD algorithm to work with any number of inputs in the input layer.
In this tutorial, which is the Part 1 of the series, we are going to make a worm start by implementing the GD for just a specific ANN architecture in which there is an input layer with 1 input and an output layer with 1 output.
In this post, we take a look at a problem that plagues training of neural networks, pathological curvature.
