Machine Learning 101: Exploring The Essence Of Supervised, Unsupervised, And Reinforcement Learning

What is Machine Learning?

What is meant when we talk about “machine learning”? Let’s go over a few ML basic concepts. Artificial intelligence plays a part in machine learning, and it enables systems to learn from huge amounts of data and solve specific problems. This makes use of computer algorithms that evolve into stronger by the time. Machine Learning is the process of programming computers to improve a series of tasks, Using sample data or prior knowledge.

Machine learning has basically 3 types:

Supervised learning, Unsupervised learning, and Reinforcement learning:

Let’s talk about what all this types are and how they are being used in our daily lifestyle – 

1: Supervised Learning:

In supervised learning, the trainer provides the system with the answer to the question together with the data used for training, allowing the machine to learn the concept and think the right output in the future.

However, does that appear that we must address every query that will be brought up? The answer is ‘No’. When we teach a machine how to recognize certain terms, it records the instructions we provide and uses them to react in the future.

We may take the alarm clock in our everyday lives as an example. When we scheduled an alarm for a particular time and day, the alarm clock remembers our input and will provide output in the future according to those settings (for example, an alarm set for each Wednesday).

Let’s use the Gmail app as another example. We occasionally mark some emails as important or spam, and that marking helps the app predict future emails.

In above image example data is labelled which makes it easy for machine to predict accurate outputs.

Supervised learning consists of Classification and Regression:

The purpose of classification algorithms is to classify or predict distinct values. such as truthful or false, male, or female, spam or not spam, etc.


For predicting constant figures like price, age, income, etc., regression algorithms are applied.

2. Unsupervised Learning:

In unsupervised learning, a machine is given an input and is then used to predict the output using data that is supplied but is not labelled. If the data does not share any similarities, this could lead to inappropriate outputs. Unsupervised learning is a subcategory of machine learning in which examples are taught using datasets that are unlabelled and are free to act on the data without being checked by a human observer.

Unsupervised learning helps in identifying of interesting facts in the data. Unsupervised machine learning development is much like how a human learns to reason through their own experiences, which brings it closer to actual artificial intelligence. Unsupervised learning is ever more important because it uses unlabeled and uncategorized data.

In real-world, we do not always have input data with the corresponding output so to solve such cases, we need unsupervised learning.

It is the machine learning task to find patterns and structures in unlabelled data set.

When humans engage captcha, they use matching patterns, which are fairly similar to those used by AI in unsupervised learning.

3. Reinforcement Learning:

In simplest terms, we can say that the state of the current input determines the output, and the result of the previous input determines the state of the current input. Because decisions in reinforcement learning are dependent, we name the sequences of these decisions.

Example: Chess game

Reinforcement learning includes feedback to better future decisions; it learns from existing failures by applying feedback.

Comment below with your real-world examples of all three forms of machine learning.

Credit – Juveria Dalvi

Leave a Reply

Your email address will not be published. Required fields are marked *

Send us your resume