What is Machine learning ? | Supervised, Unsupervised and Reinforcement learning Definition

What is Machine Learning?

Machine learning is a programme that gives computers the ability to learn without being programmed.

Machine learning gives the ability for computer programmes to analyze data, extract information automatically and, learn from it.

So basically, implementing machine learning means creating algorithms which can learn and make a prediction on data.

let's see some example of machine learning. 

• Amazon recommendation based on the customer's browsing and purchasing behavior.
• Google's search engine, that ranks the websites by relevancy.
• self-driving cars.
• Email spam filters.

All of the systems above learn themselves as more and more data is provided.

like the more you search on Amazon, it will give you the better the recommendation,

which emails you mark as a spam, the system filters new emails in a better way.

 And in the self-driving cars, the negative situation make it more accurate.

This is very similar to how we humans learn with a problem in various situation . for example, the email spam filter needs to indicate the spam and not spam emails. Note, that this is quite user dependent.

Types of machine learning 

Machine learning algorithms can be divided into three major categories.

Supervised learning

The computer is presented with examples of input and their desired outputs, and the goals are to learn a general rule that maps inputs to an output. An example is an email spam filter.

Unsupervised learning

No labels are given to the learning algorithms, leaving it on its own to find structure in its input (discovering hidden pattern data).

For example, Imagine that having all data of cars and their buyers, that is why the system can find the pattern and identify that.

 let's see another example, imagine in particular town people like small electric cars, knowing this can help system predict who will buy which car.

Reinforcement learning

A computer program interacts with the adynamic environment in which it must perform a certain goal(such as driving a vehicle or playing a game against an opponent). The program is provided with feedback in term of reward and punishment as it navigates its problems space.

Considering the desired output another classification of machine learning works is raised.

In classification, (typically in supervised learning) inputs are divided into two or more classic. Spam filtering is an example of classification, where the inputs are emails and classes are 'spam' and 'not spam'.

In regression, also a supervised problem, we predict continuously-valued outputs. for example predicting-house prices or stock prices.

In clustering, a set of inputs is divided into groups. Where in classification is nor divided like clustering, the groups are not known before, making this an unsupervised task. An example is customer segmentation.

Density estimation

Finds the distribution of inputs in some space. for example, having a diabetes test result of a specific number of people, we can estimate the distribution for the whole population 

Dimensionality reduction

Simplify inputs by mapping them into a lower-dimensional space. topic modeling is a related problem where a program is given a list of human language documents and is tasked to find out which documentation cover similar topics.

Last Words,

Machine learning is a broad topic so we can not cover all things but we define the basics of machine learning. so I hope you enjoyed it. Thank you