Machine Learning in Your Browser

If you’re working in IT, you have said the magic seven letters ChatGPT at least once in the last few months and you’ll probably remember at least one news story revolving around AI. It’s probably the hype topic of the century. Today, I want to show you an often overlooked aspect. In this article, we’ll delve into a fundamental aspect of artificial intelligence: machine learning. One of the most significant topics within machine learning in recent years has been the renaissance of neural networks. This powerful technology has proven itself to be versatile problem solver – a Swiss Army knife in the world of machine learning. However, they come with a multitude of complexities, and it’s safe to say that no one can genuinely claim to fully comprehend them. Do you believe you truly understand neural networks? Let me challenge you: How do you determine if training a neural network converges? How many training samples are required? Which training data does your network prioritise, and which does it disregard? I assume that you don’t have answers to these questions. And that’s perfectly fine! Neural networks remain an incredibly useful tool, one that can be effectively harnessed through empirical methods, even if we don’t fully understand their complexities. Amidst the dazzling allure of neural networks, unfortunately, other methods often get overshadowed. Today, I’d like to introduce you to a highly practical one. It’s easy to grasp, straightforward to implement, and incredibly versatile. However, let me make it clear from the outset: it doesn’t rival the sophistication of neural networks, but it doesn’t have to. There are many problems that can be elegantly solved with this method, offering a simple yet effective alternative to more complex solutions. <a href="https://medium.com/@franz.bender/machine-learning-in-your-browser-c8c85f8f65d3">Website</a>