PCA: orthogonal linear transform. the greatest variance by any projection of the data comes to lie on the first coordinate, the second greatest on the second coordinate, and so on
FLD: linear discriminant analysis
maximizes the distinctness between classes of data
approximation(PCA, max. variance) vs. classification (LDA)
AdaBoost
์ฌ๋ฌ๊ฐ์ weak classifier(vertical/horizontal half plane)๋ฅผํฉ์ณํ๋์ final classifier๋ฅผ๋ง๋ฌ
Cascaded Classifier
10๋ฐฐ๋น ๋ฆ - only positive examples are considered์ด๊ธฐ๋๋ฌธ
13 Neural Network
algorithms that try to mimic the brain
activation functions
sigmoid, ReLU, Leaky ReLU, tanh, Maxout, ELU
sigmoid: ๋ฏธ๋ถ๊ฐ๋ฅ, x์ domain์ด๋ฌดํํจํ์ง๋ง vanishing gradient problem ๋ฐ์
finding a free chunk of memory that can satisfy the request and splitting it into two: when request for memory allocation is smaller than the size of free chunks
The Turing test is a test, developed by Alan Turing in 1950, of a machine's ability to exhibit intelligent behaviour equivalent to, or indistinguishable from, that of a human.
Unintelligent human behavior - ๊ธฐ๋๊ฐ์ด -์์๋๋ถ๊ตฌํ๊ณ ๋ณต๊ถ์์ฌ๋ํ์
gb-f search๊ฐ์๋นํจ์จ์ ์ด๋๋ฉด์ฌํ์จ๊ฑฐ์๋ํด์๋๋ค๊น๋จน๊ธฐ๋๋ฌธ. ๋ฐ๋ผ์๊ทธ๊ฒ๋๊ณ์ฐ์ํฌํจํ๋ฉด๊ฝคํจ์จ์ ์ธ์๊ณ ๋ฆฌ์ฆ์์ฐพ์์์์๊ฒ์ด๋ค๊ทธ๊ฒ๋ฐ๋ก a*: f = g + h
17์ชฝ: at least one์ด๋ผ๋์๋ฏธ๋๋ชจ๋ partition์์๋ฏธ๋ฌ์ด๋ฉด์ ์ฒด์์๋๋ฏธ๋ฌ์ด๋ผ๋์๋ฏธ. ๋ฐ๋ผ์์ ์ด๋ํ๋์์๋ local minimum freq.๋ฅผ๋์ด์ผํจ
cosine๋ total number of transaction์์ํฅ์๋ฐ์ง์๋๋ค๋์์๊ฐ์๋ค. lift๋๋ฐ์์๋ํ๋ฉด๋ถ๋ชจ์์ํ๋์ฌ๋ผ๊ฐ์๋ถ์์๊ณฑํด์ง๋๊น
์ฑ ์์๋ฌธ์ ๋์ฌ์๋์๋ค
constraint์์ข ๋ฅ: knowledge type, data constraint, dimension, interestingness