In this series of posts, I will show how to build an image classifier using your own dataset. We’ll discuss how to prepare a dataset for machine learning, how to build a model to classify it, and techniques to improve the model’s performance. This post will focus on the data preparation.
This post shows some of the various tools in Python for visualizing images. There are usually two steps to the visualization process. First, you’ll need to read in the image from a file path, usually as a numpy array or something similar. Then, you can visualize it with various libraries.
Let’s look at how to explore images in Python. We’ll use the popular and active Pillow fork of PIL, the Python Imaging Library.
Is there sarcasm on the Internet? OK, that’s an easy one. Here’s a more difficult question: can an AI be trained to detect that sarcasm? The first step to training that algorithm would be to create a corpus of sarcastic statements from the Internet. Fortunately, there’s a lot of sarcasm out there and, even more fortunately, much of it is already labeled.
What goes on inside the black box of a machine learning algorithm? While it may be impossible for a human to understand precisely why a large neural network produced the results it did, some algorithms are far more transparent. Decision trees are just such an example of a machine learning algorithm whose results can be understood by people.
The following is my handy post for vim notes, tips, and tricks.
The story of rabbits in Australia, and the resulting eradication efforts, provides a cautionary tale about viruses and immunity. This post will explore the growth of that population and the government’s response to it.