[ML] Basic Concepts

✔︎ Introduction to ML

Maching Learning is particularly valuable for problems that  you can't solve by figuring the rules out for ourselves.

 

- A mathematical science에서 a natural science로의 전환

- 양자의 차이점은 후자가 전자보다 더 많은 예외를 허용한다는 것

 

 

✔︎ Framing

What is (supervised) machine learning?

ML systems learn how to combine input to produce useful predictions on never-before-seen data.

 

Labels

A label is the thing we're predicting—the y variable in simple linear regression.

ex) the future price of wheat, the kind of animal shown in a picture, the meaning of an audio clip, ...

 

Features

• A feature is an input variable—the x variable in simple linear regression.
• A more sophisticated machine learning project could use millions of features, specified as:

Examples

An example is a particular instance of data, x. (We put x in boldface to indicate that it is a vector.) We break examples into two categories:

• labeled examples
• unlabeled examples

A labeled example includes both feature(s) and the label. That is:

 

labeled examples: {features, label}: (x, y)

 

Use labeled examples to train the model. 

Once we've trained our model with labeled examples, we use that model to predict the label on unlabeled examples.

 

Models

A model defines the relationship between features and label.

• Training means creating or learning the model. That is, you show the model labeled examples and enable the model to gradually learn the relationships between features and label.

• Inference(추론) means applying the trained model to unlabeled examples. That is, you use the trained model to make useful predictions (y'). For example, during inference, you can predict medianHouseValue for new unlabeled examples.

 

Regression(회귀) vs. classification

- A regression model predicts continuous values.

• What is the value of a house in California?

• What is the probability that a user will click on this ad?

 

- A classification model predicts discrete values.

• Is a given email message spam or not spam?

• Is this an image of a dog, a cat, or a hamster?

 

🔑 정리하기

• label은 output, feature는 input.
• examples는 a particular instance of data. (x, y)의 묶음. labeled example인 (features, label)는 모델을 train하기 위해 사용되고, trained model에 unlated example인 (features, ?)를 대입하여 y값을 예측해낸다.

 

🏷 developers.google.com/machine-learning/crash-course/ml-intro