Supervised Machine Learning - Classification #ai #machinelearning #supervisedlearning

20 May 2025 · 1 min

Watch on YouTube

0:51

The speaker explains supervised learning using a concrete example of training a model to recognize fruit in images. By presenting thousands of labeled fruit pictures—each annotated with the correct fruit name—the model learns to differentiate fruits. The speaker emphasizes the need for diverse examples (varying colors, lighting, orientation, backgrounds) to ensure robust learning. Initially, the model makes mistakes (e.g., confusing an apple with an orange), but through iterative training it refines its predictions and ultimately achieves high accuracy in identifying the correct fruit.

Chapters

Let's imagine that we want to build a machine learning model able to identify the fruit in a picture. To do this, we show it thousands and thousands of pictures, and at the same time tell it exactly which fruit is on each of them. In this case, we're talking about supervised learning, because the datasets, which in our case are the fruit pictures, are labeled. So, in order that the model can identify the apples, we need to show it a large number of pictures of apples in different colors, with different lighting, orientation, orientations, backgrounds, and so on. At first, the model will be unable to identify the exact fruit, taking an apple for an orange, for example. But as it's trained, it will learn to refine its predictions and finally identify the fruit in question with a high degree of accuracy.