Types of Machine Learning

1. Supervised Learning          
2. Unsupervised Learning
3. Reinforcement                
4. Generative AI

Supervised Learning: Make predictions after seeing lots of data with the correct answer, and then discover the connection between the elements.
The two most common cases of Supervised learning:
  a. Regression                                   b. Classification

a. Regression: A Regression model predicts a numeric value.
   Ex: 4 Weather model that predicts the amount of rain
       4 Future house price, 4 Future ride time.

b. Classification: A classification model predicts the likelihood that something belongs to a category. Classification model outputs a value that states whether or not belong to a particular category.
   Ex: 4 Email is spam if a photo contains a cat.
   Classification models are divided into two groups:
   - Binary: Output value from a class that contains only two values.
   - Multiclass: Output value from a class that contains more than two values.

Unsupervised Learning:
An unsupervised learning model makes predictions by being given data that does not contain any correct answers. An unsupervised learning model's goal is to identify meaningful patterns among the data. A commonly used unsupervised learning model employs a technique called Clustering.

Clustering: Clustering differs from classification because the categories aren't defined by you.
Ex: An unsupervised model might cluster a weather dataset based on temperature, revealing segmentations that define the seasons. You might then attempt to name those clusters based on an understanding of the datasets.

Left — before clustering (unsupervised view). The model receives only raw feature measurements — temperature and humidity — with no labels attached. All points look identical. The structure is hidden in the data, waiting to be discovered.

Right — after clustering. The algorithm (e.g. k-means, or a classifier trained on labelled data) has grouped points by similarity. Each shape and colour encodes a class:

• Circle (blue) — Snow: low temperature, high humidity
• Square (teal) — Rain: high temperature, high humidity
• Diamond (purple) — Sleet: low temperature, mid humidity
• Triangle (amber) — No Rain: high temperature, low humidity

The dashed lines are decision boundaries — the frontiers the model learned to separate classes. Any new data point that falls into a region gets assigned that region's label.

This is the core idea behind classification in supervised ML: learn a function that maps feature space (temperature, humidity) to a discrete output class (weather type). The boundaries can be linear (like here), or highly curved for complex datasets using algorithms like SVMs with RBF kernels or neural networks.

Reinforcement Learning: It trains an agent to make decisions by interacting with an environment instead of being told the correct answer. The agent learns by the trial and error method and gets a reward for good actions and penalties for bad ones. This approach is good for problem having sequential decision making.
Such as: Robotics, Gaming, Autonomous Systems.

Generative AI: Generative AI creates new content such as text, images, and audio from a set of data.
Ex: 4 Generative AI creates unique images, music compositions, and jokes. It can summarise articles, explain how to perform a task, and edit a photo.
  text to text, text to video, text to speech
  text to image, text to code, image to text.