Thread by @_jaydeepkarale

A quick thread on types of Machine Learning Algorithms if you are just getting started.

#100Devs #100DaysOfCode

On a broad level ML algorithms are are of 4 types

1️⃣ Supervised Learning
2️⃣ Semi supervised learning
3️⃣ Unsupervised learning
4️⃣ Reinforcement

Let's explore what they mean & see further bifurcation of them

1️⃣ Supervised Learning

• A machine is taught by example
• A known dataset which includes desired inputs & outputs are provided

• The algo needs to find patterns in the input dataset, learns from observations & makes prediction with high acchrai

Examples of of Supervised learning algos are

1. Classification
2. Regression
3. Forecasting

2️⃣ Semi Supervised Learning

• Here the algo is fed both labelled & unlabelled data

• labelled data is meaningful data whilst unlabelled data is one thag lacks information

• B y understand labelled data the algo can learn to label unlabelled data

3️⃣ Unsupervised Learning

• Here there is no fixed output/label

• The algo learns patterns from data

• The algo tries to arrange datasets into meaningful clusters/groups

Examples are Clustering & Dimension Reduction

4️⃣ Reinforcement Learning

• Here we teach the machine trial & error i.e. Feedback based learning

• The algo is provided with a set of actions, params & end valies

• These algos make a sequence of decisions in order to get maximum reward

Q-learning is a popular example

If you liked the thread, do follow me @_jaydeepkarale for more content around Python, SQL, AI/ML & Software Engineering jn general.

Also, do RT the first tweet for good karma 🤝

https://twitter.com/_jaydeepkarale/status/1543878646330314752?t=Kglw6cPpSpsvLulJ0HC-2Q&s=19