{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "\n",
    "# unit 0.4 - Types of learning\n",
    "\n",
    "There are different types of learning in machine-learning, depending on how a desired relationship is acquired.\n",
    "\n",
    "## Supervised\n",
    "\n",
    "$y_d = f(x)$ with $y_d$ as desired output and $x$ as input.\n",
    "\n",
    "Samples in the dataset are {{x1,y1}, {x2, y2}, {x3, y3}} -- both $x$ and $y$ are provided as examples!\n",
    "\n",
    "Examples:\n",
    "\n",
    "- a picture of a cat and a category number (1 = cat)\n",
    "\n",
    "![](images/cifar10.png)\n",
    "\n",
    "- from a list of homework grades (a vector $x$) to final class grade $y$\n",
    "- generating an image from a text caption (Dall-E)\n",
    "\n",
    "![](images/dalle.png)\n",
    "\n",
    "\n",
    "## Unsupervised\n",
    "\n",
    "There is only $x$ here! Can we split $x$ in groups?\n",
    "\n",
    "![](images/unsupervised.jpg)\n",
    "\n",
    "Examples:\n",
    "- a zoo of animals, all of different kinds - group by kind\n",
    "- a set of points, group by closeness (k-means clustering)\n",
    "\n",
    "![](images/kmeans.png)\n",
    "\n",
    "- a set of people, group by age, race, sex, ...\n",
    "\n",
    "\n",
    "## Self-supervised\n",
    "\n",
    "There is only $x$ here! What is the desired output $y_d$? It is $x$ itself!\n",
    "\n",
    "Examples: \n",
    "- A portion of an image from the rest of the image (in-painting)\n",
    "\n",
    "![](images/inpainting.png)\n",
    "\n",
    "- Removing noise from an image (de-noising)\n",
    "\n",
    "![](images/denoising.jpg)\n",
    "\n",
    "- predicting the next words from a previous set of words (language models like: chat-GPT, Gemini, Llama)\n",
    "\n",
    "![](images/predictwords.png)\n",
    "\n",
    "## Reinforcement learning\n",
    "\n",
    "Learning to improve a task based on a performance metric. Self-improvement by comparing your results and trying to optimize gains. \n",
    "\n",
    "![](images/rl.png)\n",
    "\n",
    "Examples:\n",
    "- a program that learns to play chess, go, tic-tac-toe, etc. from playing the game\n",
    "- an athlete that wants to improve its moves to achieve better results\n",
    "- a robot that learns to pick-and-place objects by trying not to drop them, or improving speed\n",
    "\n",
    "\n",
    "## Imitation learning\n",
    "\n",
    "Copy a sequence of items from one instance to another. Learning to do something by observing someone else.\n",
    "\n",
    "![](images/il.png)\n",
    "\n",
    "Examples:\n",
    "- learning to draw a figure by observing someone\n",
    "- learning to play music from a youtube video\n",
    "- learning a recipe from a video\n",
    "- learning to code machine-learning from video-lectures\n",
    "- learning to sing after someone\n",
    "\n",
    "![](images/robot-imitation.jpg)\n",
    "\n",
    "\n",
    "## Summary\n",
    "\n",
    "These are the most typical learning techniques and data samples in the respective datasets. The examples of applications are innumerable and we only mentioned a few here for brevity.\n",
    "\n",
    "\n",
    "## References\n",
    "\n",
    "Just a few links here but there are many more. Search online for \"types of learning\".\n",
    "\n",
    "[1-types of learning](https://www.coursera.org/articles/types-of-machine-learning)\n",
    "\n",
    "[2-types of learning](https://machinelearningmastery.com/types-of-learning-in-machine-learning/)"
   ]
  }
 ],
 "metadata": {
  "language_info": {
   "name": "python"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}