How AI Learns with Labeled Data | Supervised Learning

Key Takeaways

Hey folks, I'm back with new video. So in this video, we'll discuss about uh supervised learning. Well, supervised learning is a part of a statistical machine learning. So let me take an example to explain u supervised learning more detail. Uh imagine I do have a student uh you can consider him a kid 5 year old. I'll be the teacher for this kid. My job is to teach following shapes to the kid. I do have shapes and their names. Slowly I'll show the shape and I'll tell the name to the kid and kid will start learning shape and name. Now I'll take a new shape and I'll ask the kid whether he got the proper understanding on the topic. So we have two scenarios here. If student predicted the right output then the kid is absolutely good. So he got the complete concept. But what if student um got the output as wrong? If I show a circle image, if he predicted as a rectangle, then I need to go back to the kid. I need to change my learning pattern and again I need to make him to understand the shapes. Until he predicts the right output, I'll do this process. Now let's look this entire scenario in supervised learning perspective. Uh let's take this kid as our model. Now we are training this model with input label data. This shapes are data. This names are labels. We are training the model with label data. Now I will take a new data and validate whether the model is perfectly good enough. Now if I show a rectangle shape and I'll ask the model to predict. If it predicts as rectangle then the model is absolutely good. If it predicts the rectangle shape as circle then I'll train the model again and again until it predicts the right output. So in a supervised learning you need to remember things like uh we send label data to the model. model will predict and give us the output as label. And supervised learning has two types regression and classification. We'll discuss about regression and classification in coming videos. Thank you for watching. Uh signing off Mittton.