Luckily, studying for a degree in mathematics is very different from high school.
Math class didn’t exactly get the best reputation in school — and I can see why. It tells students to learn about seemingly pointless techniques and memorise different formulas that 99% of them will never use again… unless they go on to become math teachers.
Things change, however, when you decide to study maths at a higher level.
This is exciting! We’re nearly ready to start simulating a whole game of Chess. There’s just a few more components we need.
To begin, let’s first consider what information we need to encode to fully describe a move.
In standard chess move notation, we are given only the type of piece, and it’s new location. E.g. Ne2 which indicates that a knight should be moved to the square e2.
If you haven’t already heard, Strava provide a powerful API that allows users to access their data using code. This gives us the ability to create interesting visualisations, analyse progress, and create custom trackers and planners, all with just a few lines of code.
This is part 2 in a series of articles documenting my process of building a chess engine. I hope to share insights from my previous experience, and explain some of the concepts and techniques I’ll be using. If you haven’t already read part 1, I’d strongly recommend taking a look before reading part 2. You can check it out at the link above.
Time to pick up where we left of…
Looking at the images above, we see an example of an image posterization filter that gives images a cartoon-like appearance, but behind the scenes, this filter is actually using a machine learning algorithm known as clustering.
Before exploring into how this process works and seeing how we can implement it in Python, let’s take a look at why we might want to do this in the first place.
This is not my first rodeo.
I’ve built a chess engine before, but the results were… let’s say… underwhelming.
Don’t get me wrong, the program worked; it was able to play a full game of chess, and for that, I give it credit. But let’s face it, it was terrible at chess.
The Travelling Salesman Problem, TSP, describes a scenario where a salesman wishes to visit a number of cities, while taking the shortest possible route, before returning home to the start point. While it may appear simple, this problem not only has no known polynomial time solution, but there is also no time-efficient way to prove that a given solution is in fact optimal.
Imagine you want to travel around Europe. You want to see the sights, and visit as many European capital cities as possible, but you’re on a budget; flights can get expensive, so you plan to travel by rail and bus. To make your funds stretch as far as possible, you need to plan a route that will take you to all of the cities on your list for the cheapest price, before heading home again.
In the ever-expanding field of artificial intelligence, some of the most fascinating advancements come from mimicking the natural world. One such breakthrough is the Genetic Algorithm, a powerful machine learning technique inspired by Charles Darwin’s theory of natural selection. Just as organisms evolve over time, honing their adaptations to thrive in their environments, Genetic Algorithms evolve solutions to complex problems, becoming increasingly effective with each iteration.
