- Foundation
- About 15 min
- 5 practice questions
Branches of AI: machine learning, deep learning, NLP, computer vision, speech and robotics
Name the main branches of AI, say what each does with an everyday example, and explain how AI, machine learning and deep learning fit inside one another.
If something looks wrong, please tell us.
AI is a large field with several branches. Questions often give a situation - a phone reading a number plate, an app translating Hindi into English - and ask which branch is at work. This lesson gives you a map of the branches and a quick way to identify each.
AI, machine learning and deep learning: circles inside circles
Picture three circles, one inside the next. AI is the biggest circle: any technique that makes machines perform intelligent tasks. Inside it is machine learning (ML): AI that learns from data instead of following only hand-written rules. Inside ML is deep learning (DL): machine learning that uses neural networks with many layers. So all deep learning is machine learning, and all machine learning is AI, but not the other way round.
Neural networks in one paragraph
An artificial neural network is loosely inspired by the brain's neurons. It has an input layer that receives data (such as the pixels of a photo), one or more hidden layers that process it, and an output layer that gives the answer (such as 'cat'). Each connection has a weight that is adjusted during training. 'Deep' simply means many hidden layers, which lets the network learn complex patterns step by step - edges, then shapes, then whole objects.
Layers of a neural network
- Input layer Receives the data, such as the pixels of a photo.
- Hidden layers Process it; many hidden layers make a network deep.
- Output layer Gives the answer, such as 'cat'.
Three ways machines learn
Three ways machines learn
-
Supervised learning
From examples labelled with the right answer.
- Emails marked spam or not spam
-
Unsupervised learning
From unlabelled data, finding groups or patterns on its own.
- Grouping shoppers with similar buying habits
-
Reinforcement learning
By trial and error, with rewards and penalties.
- A program learning to play a game by playing it many times
| Type of machine learning | How it learns | Example |
|---|---|---|
| Supervised learning | From examples labelled with the right answer | Emails marked spam or not spam |
| Unsupervised learning | From unlabelled data, finding groups or patterns on its own | Grouping shoppers with similar buying habits |
| Reinforcement learning | By trial and error, earning rewards for good actions and penalties for bad ones | A program learning to play a game by playing it many times |
The main branches
| Branch | What it does | Everyday example |
|---|---|---|
| Machine learning | Learns patterns from data to predict or decide | Fraud alerts on unusual payments, product recommendations |
| Deep learning and neural networks | Machine learning with many-layered networks for complex data | Face recognition, modern speech recognition |
| Natural language processing (NLP) | Understands and produces human language | Hindi-English translation, chatbots, grammar suggestions |
| Computer vision | Understands images and video | Face unlock, number-plate reading, spotting crop disease in leaf photos |
| Speech recognition | Turns spoken words into text | Voice typing, voice assistants |
| Robotics | Machines that sense and act in the physical world | Warehouse robots, robot vacuum cleaners, factory robot arms |
| Expert systems | Apply if-then rules taken from human experts | Early medical-advice systems |
Two cautions. First, a robot is not automatically AI: an arm that repeats the same welding motion all day is automation. It becomes AI when it senses and decides, for example a warehouse robot that finds its way around obstacles. Second, real products usually combine branches, as the first worked example shows.
Worked example 1: one voice question, four branches
- You say, 'Kal Jaipur mein mausam kaisa rahega?' (What will the weather be in Jaipur tomorrow?)
- Speech recognition converts your voice into text.
- NLP works out the meaning: a weather request, for Jaipur, for tomorrow.
- The assistant fetches the forecast from a weather service.
- Speech synthesis (text-to-speech) reads the reply aloud. Deep learning powers most of these steps.
Worked example 2: name the branch
| Situation | Branch | Clue |
|---|---|---|
| A toll plaza camera reads vehicle number plates | Computer vision | Understanding an image |
| An app translates a Tamil sign into English | NLP (with computer vision to read the sign) | Language |
| A bank groups customers by spending habits without labels | Machine learning (unsupervised) | Finding groups in unlabelled data |
| A farmer speaks a query and it appears as text | Speech recognition | Voice to text |
AI ⊃ machine learning ⊃ deep learning. Language → NLP; images → computer vision; voice to text → speech recognition; physical action → robotics; if-then expert rules → expert systems.
Key takeaways: deep learning sits inside machine learning, which sits inside AI. Neural networks have input, hidden and output layers; 'deep' means many hidden layers. ML learns in supervised, unsupervised or reinforcement ways. Identify a branch by its input: language, images, speech, physical action or expert rules.
Check your understanding
-
1. Which statement about AI, machine learning (ML) and deep learning (DL) is correct?
- DL is a part of ML, and ML is a part of AI
- AI is a part of ML
- ML is a part of DL
- The three are unrelated fields
Show answer
Answer: A. Think of circles inside circles: AI is the largest, ML is AI that learns from data, and DL is ML that uses many-layered neural networks. So DL lies inside ML, which lies inside AI.
-
2. An app reads the printed text on a photographed electricity bill and fills a payment form. Reading text from the image mainly uses:
- Reinforcement learning
- Speech recognition
- Computer vision
- Robotics
Show answer
Answer: C. Getting information out of an image is computer vision (reading printed text from images is often called OCR, optical character recognition). No speech, physical movement or reward-based learning is involved.
-
3. A program learns to play a game by playing it repeatedly, getting points for good moves and losing points for bad ones. This is:
- Supervised learning
- Reinforcement learning
- An expert system
- Unsupervised learning
Show answer
Answer: B. Learning by trial and error with rewards and penalties is reinforcement learning. Supervised learning needs labelled examples, and unsupervised learning finds groups in unlabelled data.
-
4. In a neural network, the word 'deep' refers to:
- How expensive the computer is
- The network being stored underground
- The network having only one layer
- The network having many hidden layers
Show answer
Answer: D. Deep learning uses neural networks with many hidden layers between the input and output layers. More layers let the network learn complex patterns step by step.
-
5. Which branch of AI deals with understanding and producing human language, as in chatbots and translation?
- Natural language processing (NLP)
- Computer vision
- Robotics
- Expert systems
Show answer
Answer: A. NLP is the branch for human language: translation, chatbots, summaries and grammar suggestions. Computer vision handles images, robotics handles physical action, and expert systems apply if-then rules.
Sign in to keep track of the lessons you finish.