Cousera机器学习基石第五周笔记-Machine-Learning-Foundation-Week-5-Note-in-Cousera

1 min

Traning versus Testing

Actually,I didn’t fully understand this part of course.However I don’t focus on the theory but application.

I will update this note if possible.

Recap and Preview

Recap: The ‘Statistical’ Learning Flow

Two Central Questions

1. can we make sure that Eout(g)E_{out}(g) is close enough to Ein(g)E_{in}(g)?

2. can we make Ein(g)E_{in}(g) small enough?

\Rightarrow What role does MH\underbrace{M}\\{|\mathscr{H}|} play for the two questions?

## Trade-off on *M

small Mlarge M
1. Yes P[BAD]2Mexp(...)\mathbb{P}[BAD]\leq2\cdot M\cdot exp(...)1. No!P[BAD]2Mexp(...)\mathbb{P}[BAD]\leq2\cdot M\cdot exp(...)
2. No!\color{Maroon}No!,two few choices2. Yes!many choices

using the right M is important

Preview

  • Know: P[Ein(g)Eout(g)>ϵ]2Mexp(2ϵ2N)\mathbb{P}[|E_{in}(g)-E_{out}(g)|>\epsilon]\leq 2 \cdot M\cdot exp(-2\epsilon^2N)
  • To do
    • establish a finite quantity that replaces MP[Ein(g)Eout(g)>ϵ]2mHexp(2ϵ2N)\mathbb{P}[|E_{in}(g)-E_{out}(g)|>\epsilon]\leq 2 \cdot m_{\mathscr{H}}\cdot exp(-2\epsilon^2N)
    • justify the feasibility of learning for infinite M
    • study mHm_{\mathscr{H}}to understand its trade-off for ‘right’H\mathscr{H},just like M

Effective Number of Lines

Where Did M Come From

The BAD event :uniform bound fail\rightarrow M\approx\infty

Where Did Uniform Bound Fail

overlapping for similar hypotheses

union bound over-estimating

To acoount for overlap,can we group similar hypotheses by kind?

How Many Lines Are There?

In 2D,we can get smaller than 2N2^Nlines

Effective Number of Lines

maximun kind of lines with respect to N inputs x1x_1,x2x_2,…,xNx_N

    \iff==effctive number of lines==

In 2D,the effective number of lines

  • must be 2N\leq 2^N

  • finite ‘grouping’ of infinitely-many lines H\in H

  • wish:

    P[Ein(g)Eout(g)>ϵ]2effctive(N)exp(2ϵ2N)\mathbb{P}[|E_{in}(g)-E_{out}(g)|>\epsilon]\leq 2 \cdot effctive(N)\cdot exp(-2\epsilon^2N)

If

  1. effective(N) can replace M and

  2. effective(N) 2N\ll 2^N

    ==learning possible with infinitie lines==

Effective Number of Hypotheses

Dichotomies

Break Points