The Learning Problems

The Learning Problems

  • When Can Machines Learn? (illustrative + technical)
  • Why Can Machines Learn? (theoretical + illustrative)
  • How Can Machines Learn? (technical + practical)
  • How Can Machines Learn Better? (practical + theoretical)

When Can Machines Learn?

Some definitions

Learning: acquiring skill with experience accumulated from observations

Machine learning: acquiring skill with experience accumulated/computed from data

data ——-> ML ———> sklill

Skill: improve some performance measure (e.g. prediction accuracy)

data ——-> ML ———> improved performance measure

The Machine Learning Route

ML: an alternative route to build complicated systems

Give a computer a fish, you feed it for a day; teach it how to fish, you feed it for a lifetime

Key Essence of Machine Learning

Improving some performance measure with experience computed from data

  • exists some underlying pattern to be learned
  • but no programmable (easy) definition
  • somehow there is data about the pattern

Formalize the Learning Problem

  • inputs: $x \in \chi$
  • outputs:$y\in \gamma$

  • unknown pattern to be learned <=> target function

  • data <=> training examples

  • hypothesis <=> skill with hopefully good performance

Learning Flow

Learning Model.png

Learning Model

Learning Flow.png

Practical Definition of Machine Learning

Machine Learning: use data to compute hypothesis g that approximates target f

Relationships between Machine Learnig, Data Mining, Artifical Intelligence and Statics

  • Machine Learning

use data to compute hypothesis g that approximates target f. ML can realize AI, among other routes($g \approx f$ is something that shows intelligent behavior)

  • Data Mining

use (huge) data to find property that is interesting.

  • Artifical Intelligence

compute something that shows intelligent behavior.

  • Statics

use data to make inference about an unknown process, statistics is a useful tool for ML.