Ocam's Razor: If there are multiple hypotheses that are consistent with observations, choose the simplest one.

Important premise of NFL theorem

This set of records

Each record is a description of an event or object

Matters that reflect the performance or nature of an event or object in some aspect

Since each point in space corresponds to a coordinate vector, we also call an example an "Feature vector"

Generally, let D={x1,x2,..,xm} represent a data set containing m examples, each example is described by d attributes (for example, the watermelon data above uses 3 attributes), then each example xi=(xi1;xi2;...xid) is a vector in the d-dimensional sample space The value on the attribute is "stiffness"), and d is called the "dimensionality" of sample xi.

The process of learning a model from data is called "learning" or "training". This process is completed by executing a learning algorithm.

The learned model corresponds to a certain potential law about the data, so it is also called a "hypothesis"; this potential law itself is called "truth" or "reality", and the learning process is to find out or approach the truth.

"Supervised learning"

"Unsupervised learning"

"Semi-supervised learning"

The goal of machine learning is to make the learned model well applicable to "new models". ", rather than just the training set, we call the model's ability to adapt to new samples Strength is the ability of generalization.

Induction, generalization from the specific to the general process.

deduction, the process from general to "specialization". Machine learning is essentially inductive learning.

A hypothesis formed from all the values ​​that the attribute can take.

A "set of hypotheses" consistent with the training set.