Reinforcement Learning

Machine Learning

Machine Learning or Statistical Learning as the name suggest is a branch of Artificial Intelligence algorithms that learn specific tasks from data. Machine Learning algorithms can be devided into three categories:

• Supervised Learning relies on annotated data.
• Unsupervised Learning consists of finding pattern in unlabled data.
• Reinforcement Learning algorithms on the other hand learn from data they generated by interacting with an environment.

  Reinforcement Learning

  Imagine you’re scrolling Instagram as one does at 1 AM and you come across food posts - as one does - and you start wondering. “Wouldn’t it be great if I could recreate those scrumptious meals right now!”, and if you’re anything like me (undiagnosed but something is definetely wrong (•ᴗ•،،)) you’d start thinking about building a robot chef. Let’s call our chef Jarvis. Jarvis being a robot needs a recipe to be able to recreate a dish. However when you go under your favorite food influencer on instagram and you check the descriptions on his posts for the recipes, all that you find is a link to buy their cookbook. Determined to finish you project you start looking for ways to generate recipes based on the picture of the dishes. You first look into the most straight forward solution which is to learn a model by using annotated (with the recipe) data (picture of dishes), but you quickly come to the realisation that the lack of annotated data is the reason you’re building Jarvis in the first place.