The Fantastic Four of AI
The Layman’s Guide to Machine Learning/Deep Learning Lies Here
Machine learning is the bag of techniques that you can use to teach a machine to do some task like it is a dog or cat (classification) or what will be the temperature tomorrow (regression). Deep learning is doing machine learning but sticking specifically to using brain-like structures i.e. neural networks. At the core of it lies our inability today to analyze the big data we generate on a daily basis. In short, we need the machines to do our work. We are good at finding patterns in things, we just want the same from machines.
There are four types of learning.
- Supervised Learning
- Unsupervised Learning
- Semi-supervised Learning
- Reinforcement Learning
Supervised Learning
Focus on the word ‘supervision’. You are going to be supervised but by what? Labels.
On one hand, you have data and on the other, you have labels.
In the supervised setting, the ML/DL algorithm improves by checking against the labels (or ground truth) it has. The more predictions match with the labels the better is (up to a point).
