Skip to main content

Machine Learning Landscape

Artificial Intelligence is the broad umbrella. Underneath it are the actual active branches of software development that power modern systems.

Branches of AI​

  • Machine Learning (ML): Teaching a machine to learn from data without explicit programming.
  • Deep Learning (DL): A subset of ML utilizing massive multi-layered neural networks.
  • Natural Language Processing (NLP): Giving machines the ability to read, understand, and generate human language text.
  • Computer Vision: Training algorithms to understand the visual world (images and videos).
  • Robotics: Integrating AI into physical machinery to navigate and act upon physical environments.

Machine Learning Basics​

Machine Learning shifts the programming paradigm entirely. Instead of writing rigid if-then rules, you provide a model with data and answers, and it mathematically figures out the rules itself.

The Three Learning Paradigms​

1. Supervised Learning​

The model is trained on a labeled dataset. You provide both the input (features) and the desired output (labels).

  • Use Case: Predicting house prices. You feed the model 10,000 houses with their square footage and their final sale price. The model learns the correlation.

2. Unsupervised Learning​

The model is given unlabeled data. The system tries to learn the patterns and structures completely on its own without a defined "correct answer".

  • Use Case: Customer segmentation. You feed an e-commerce model a spreadsheet of user clicks and purchases, and it self-organizes the users into unique marketing demographic clusters.

3. Reinforcement Learning​

The model learns to navigate an environment through trial and error, aiming to maximize a mathematical "reward".

  • Use Case: An AI learning to play Super Mario. If it moves right, it gets +1 reward. If it dies, it gets -10 reward. After thousands of iterations, it learns the perfect path to win the game.