The final project involved collecting and analysing data from hundreds of Pokémon, developing a smart recommendation system to suggest similar Pokémon, using genetic algorithms to assemble the best "6-Pokémon battle team", and training a computer vision model that can automatically recognise Pokémon in images.
And as part of that - entirely voluntarily - collecting and manually labelling 1,403 Pokémon objects... all for a good result. Only downside: I now even dream about Pokémon 😵💫
Although Pokémon was initially - for some of us somewhat ...ahem older students - outside our familiar field of expertise, that was exactly our lecturer's intention: to take us out of our comfort zone and teach us to immerse ourselves in new domain knowledge. In case you did not know: Pokémon is, to say the least, an interesting and surprisingly deep domain.
AI evolves at breakneck speed, so it is important for us to keep learning. Fortunately, the teachers and lecturers at Howest do what is needed to keep the course material up to date. A special thank you to Gilles Depypere and the team at HOWEST University of Applied Sciences for organising this Masterclass!
The results will follow in a few weeks, but my AI knowledge has certainly been sharpened again. Gotta catch 'em all! 😉
