Key Components

  1. The data that we can learn from.
  2. A model of how to transform the data.
  3. A loss function that quantifies the badness of our model.
  4. An algorithm to adjust the model’s parameters to minimize the loss.


Category

Supervised Learning

Supervised learning addresses the task of predicting targets given inputs.

Regression

When our targets take on arbitrary values in some range, we call this a regression problem.
A good rule of thumb is that any How much? or How many? problem should suggest regression.
Solution: minimize the L1 or L2 loss functions

Classification

This kind of system is referred to as optical character recognition (OCR), and the kind of problem it addresses is called classification
Solution： loss function for classification problems is called cross-entropy(mushroom Problem)

Tagging

The problem of learning to predict classes that are not mutually exclusive is called multi-label classification. Auto-tagging problems are typically best described as multi-label classification problems

Search and ranking

In the field of information retrieval, we want to impose a ranking on a set of items.

Recommender system

Recommender systems are another problem setting that is related to search and ranking. The problems are similar insofar as the goal is to display a set of relevant items to the user.

Sequence Learning

It require a model to either ingest sequences of inputs or to emit sequences of outputs (or both!). These latter problems are sometimes referred to as seq2seq problems.

Unsupervised learning

Clustering

Subspace estimation problems(If the dependence is linear, it is called principal component analysis)

Representation learning and it is used to describe entities and their relations

causality and probabilistic graphical models

generative adversarial networks (GANs)

Interacting with an Environment

Above are called offline learning

Reinforcement Learning

If you are interested in using machine learning to develop an agent that interacts with an environment and takes actions, then you are probably going to wind up focusing on reinforcement learning (RL).
At each timestep t, the agent receives some observation 𝑜𝑡ot from the environment and must choose an action 𝑎𝑡at that is subsequently transmitted back to the environment via some mechanism (sometimes called an actuator).