Branches of AI - telivaina/ai GitHub Wiki
Branches of AI ๐
1. Rule-Based Systems ๐ก
AI systems that make decisions based on predefined "if-then" rules, without learning from data.
Subfields
- Expert Systems: Simulate human expertise in specialized areas like medical diagnostics.
- Fuzzy Logic Systems: Handle uncertainty by incorporating degrees of truth in decisions.
Related Fields
- Knowledge Representation: Representing knowledge in a form that a system can use to solve complex tasks.
Famous Applications
- MYCIN: A medical rule-based expert system.
- XCON: Configured computer systems for DEC.
2. Symbolic AI ๐ค
AI that represents knowledge using symbols and logical rules to mimic human-like reasoning.
Subfields
- Knowledge Representation: Forms like graphs and ontologies to represent information.
- Automated Reasoning: Drawing conclusions based on logical rules.
Related Fields
- Cognitive AI: Mimics human thinking using symbolic logic and reasoning.
Famous Applications
- IBM Watson: Uses symbolic reasoning for complex domain questions.
3. Machine Learning (ML) ๐
A subset of AI where systems learn from data to improve performance over time.
Subfields
- Supervised Learning: Learning from labeled data to make predictions.
- Unsupervised Learning: Finding patterns in unlabeled data.
- Reinforcement Learning: Learning through rewards and penalties in an environment.
Related Fields
- Deep Learning (DL): A subset of ML using neural networks with many layers.
Famous Applications
- Google: Uses ML for Search and Translate.
- Netflix: Movie recommendations.
4. Natural Language Processing (NLP) ๐ฃ
AIโs ability to process and generate human language.
Subfields
- Text Classification: Categorizing text into predefined groups.
- Machine Translation: Automatically translating text between languages.
- Sentiment Analysis: Analyzing sentiment in text (positive, negative, neutral).
Related Fields
- Computational Linguistics: Uses computational methods to process language.
- Speech Recognition: Converting speech to text.
Famous Applications
- Google Translate: Machine translation service.
- Siri / Alexa: Voice-activated assistants using NLP.
5. Computer Vision ๐๏ธ
AI that enables computers to interpret and make decisions based on visual data.
Subfields
- Image Recognition: Identifying objects or faces in images.
- Object Detection: Locating specific objects within images.
- Image Segmentation: Dividing an image into regions for analysis.
Related Fields
- Image Processing: Enhancing and manipulating images for use in vision tasks.
Famous Applications
- Tesla Autopilot: Uses computer vision for self-driving cars.
- Facebook: Uses computer vision for facial recognition in photos.
6. Robotics ๐ค
AI applied to robots, enabling them to perform tasks autonomously or semi-autonomously.
Subfields
- Autonomous Navigation: Robots navigating using sensors and data.
- Robot Perception: Allowing robots to understand their environment.
Related Fields
- Human-Robot Interaction: Focus on interactions between humans and robots.
Famous Applications
- Boston Dynamics: Known for robots like Spot.
- iRobot: Creator of Roomba, an autonomous vacuum cleaner.
7. Planning & Reasoning ๐ง
AI systems that make decisions based on goals, constraints, and environments.
Subfields
- Automated Planning: Creating a sequence of actions to achieve a goal.
- Decision Making: Making the best decision in a given situation.
Related Fields
- Optimization: Finding the best solution to a problem.
Famous Applications
- Uber: Uses automated planning to optimize routes.
- Air Traffic Control: Uses planning and reasoning for managing air traffic.
8. Deep Learning (DL) ๐
A subset of ML using neural networks with many layers to handle complex tasks.
Subfields
- Convolutional Neural Networks (CNNs): Used for image processing.
- Recurrent Neural Networks (RNNs): Used for time-series or sequential data.
- Generative Adversarial Networks (GANs): Used for generating realistic data like images.
Related Fields
- Neural Networks: Computational models inspired by the human brain.
Famous Applications
- DeepMind: Known for AlphaGo, which beat human players in the game of Go.
- OpenAI: Uses deep learning for language models like GPT.
9. Reinforcement Learning (RL) ๐ฎ
A type of machine learning where agents learn by interacting with the environment and receiving rewards or penalties.
Subfields
- Model-Free RL: Learning from experience without modeling the environment.
- Model-Based RL: Building a model of the environment to plan future actions.
Related Fields
- Game Theory: A framework for analyzing competitive interactions, often applied in RL.
Famous Applications
- DeepMindโs AlphaZero: Uses RL to play board games like Chess and Go.
- OpenAI Five: Uses RL to play Dota 2.
10. Artificial General Intelligence (AGI) ๐
AI capable of performing any intellectual task that a human can.
Subfields
- Cognitive Modeling: Simulating human thought processes in AI systems.
- Cross-Domain Intelligence: Enabling AI to generalize knowledge across various domains.
Related Fields
- Cognitive Science: Studies the mind and intelligence, relevant for AGI development.
Famous Applications
- OpenAI: Developing AGI with a focus on GPT models.
- SingularityNET: Working on decentralized AGI.
11. Generative AI (GAI) ๐ง
AI systems that create new content like images, text, or music based on learned data.
Subfields
- Generative Adversarial Networks (GANs): Models that generate new, realistic data by pitting two networks against each other.
- Variational Autoencoders (VAEs): A type of generative model used for image generation.
Related Fields
- Creative AI: AI that generates creative content such as art, music, or literature.
Famous Applications
- DeepArt: AI that creates artwork based on famous art styles.
- OpenAI DALLยทE: A model that generates images from textual descriptions.
12. Evolutionary Algorithms ๐งฌ
Optimization algorithms inspired by natural selection, using concepts like mutation and selection to evolve solutions.
Subfields
- Genetic Algorithms: Solving problems by evolving solutions through selection, mutation, and crossover.
- Differential Evolution: A variant of evolutionary algorithms that optimizes problems based on solution differences.
Related Fields
- Optimization: The process of improving solutions.
Famous Applications
- NASA: Uses evolutionary algorithms for spacecraft design and mission planning.
13. Swarm Intelligence ๐
AI based on collective behavior in decentralized systems, inspired by social insects like ants and bees.
Subfields
- Ant Colony Optimization: Mimics ant foraging behavior for optimization tasks.
- Particle Swarm Optimization: Optimizing solutions by mimicking the flocking behavior of birds.
Related Fields
- Multi-Agent Systems (MAS): Systems where multiple agents interact to solve problems.
Famous Applications
- Optimization Systems: Used for logistics and network design in companies.
14. Expert Systems ๐งโโ๏ธ
AI systems designed to emulate human expertise in specialized domains like medical diagnosis.
Subfields
- Knowledge Base: A repository of domain-specific knowledge.
- Inference Engine: Applies rules to knowledge to make decisions or conclusions.
Related Fields
- Rule-Based Systems: Foundation for expert systems that use predefined rules for decision making.
Famous Applications
- MYCIN: Medical expert system for diagnosing blood infections.
- DENDRAL: A system for chemical analysis.
15. Cognitive Computing ๐ง
AI systems designed to simulate human thought processes, including learning, understanding, and decision-making.
Subfields
- Machine Perception: Understanding sensory inputs like vision and sound.
- Natural Language Understanding: Understanding and processing human language.
Related Fields
- Artificial Intelligence: The broader field encompassing cognitive computing.
Famous Applications
- IBM Watson: A cognitive computing platform used in healthcare and finance.
- Apple Siri: Uses cognitive computing for voice recognition and interaction.