필터,filter

Difference between r1.24 and the current

@@ -1,10 +1,16 @@
#noindex
##======필터,filter =,filter 필터 filter
거름망, 거르개, 거름종이, .. 암튼 거르는 도구인데

filter가 하는 행동이 [[필터링,filtering]]? { [[신호,signal]]에 포함된 잡음이나 간섭 성분을 제거하거나, ... }
or [[여과,filtration]]?

이미지처리 ([[이미지,image]] [[처리,processing]] [[이미지처리,image_processing]])에선 tbw
----
Sub: // 이건 [[추정,estimation]]으로 묶?
Bayes_filter WpEn:Bayes_filter
Kalman_filter WpEn:Kalman_filter
----
<<tableofcontents>>

= CNN에서 filter =
@@ -82,16 +88,62 @@
[[파이썬,Python]]의 {{{filter(함수, 반복가능한것)}}}은
{{{반복가능한것}}}이 {{{함수}}}에 입력되었을 때 return_value 가 [[참,true]]인 것만 걸러내어 돌려준다? chk

= 수학의 filter =
MKLINK
= 수학(esp. order theory, lattice theory)의 filter =
번역은?
KmsE:filter 에선 그냥 필터로 함.
뜻(번역후보)는 여과 거름 .... 등등.
 
poset의 special subset, describing "large" or "eventual" elements. (we)
 
MKL
[[구조,structure]]
[[ultrafilter]] =,ultrafilter . ultrafilter
{
"ultra filter 극대필터, 초필터" via KmsE:"ultra filter"
 
주어진 poset $P$ 위의 '''ultrafilter'''는 P의 특정 subset인데,
WtEn:ultrafilter ?
[[WpEn:Ultrafilter]]
[[WpEn:Ultrafilter]] = https://en.wikipedia.org/wiki/Ultrafilter
https://artofproblemsolving.com/wiki/index.php/Ultrafilter
}

[[극대필터,maximal_filter]] =극대필터,maximal_filter =,maximal_filter . 극대필터 maximal_filter
{
'''maximal filter'''
극대필터 via kms.
 
[[극대,maximal]]
[[maximal_element]] - 극대원 or 극대원소
 
"maximal filter 극대필터
maximal filterbase 극대필터기저" via KmsE:"maximal filter"
}// maximal filter ... Naver:"maximal filter" Ggl:"maximal filter"
 
minimal_filter ?
 
[[proper_filter]] =,proper_filter =,proper_filter . proper_filter
{
'''proper filter'''
"proper filter 진필터" via KmsE:"proper filter"
} // proper filter .... Naver:"proper filter" Ggl:"proper filter" Bing:"proper filter"
 
[[prime_filter]] =,prime_filter =,prime_filter . prime_filter
{
prime filter
'''prime filter'''
Sub:
[[completely_prime_filter]]
} // prime filter .... NN:"prime filter" Naver:"prime filter" Bing:"prime filter"
 
[[completely_prime_filter]] =,completely_prime_filter =,completely_prime_filter . completely_prime_filter
{
'''completely prime filter'''
https://ncatlab.org/nlab/show/completely+prime+filter
} // completely prime filter .... Naver:"completely prime filter" Bing:"completely prime filter"
 
----
MKL
Martin_measure =,Martin_measure . Martin_measure
{
Martin measure
@@ -105,6 +157,7 @@
Up: [[측도,measure]]
}

----
comparison of filters
같은 집합에 대한 filter들간의 비교. 어떤 filter가 다른 filter보다
* finer
@@ -113,7 +166,31 @@
* strictly coarser
* comparable
에 대해: https://planetmath.org/comparisonoffilters
[[비교,comparison]]]
Up: [[비교,comparison]]]
 
----
REL
[[filtration]] =,filtration =,filtration . filtration (of mathematics)
{
'''filtration'''
 
Not in KmsE:filtration as of [[Date(2024-07-05T11:13:33)]] ....
 
}// filtration .... NN:filtration
 
----
분야는 (via we)
[[순서론,order_theory]] or 순서이론 (curr [[순서,order]])
[[격자론,lattice_theory]] or 격자이론 (curr [[래티스,lattice]] [[격자,lattice]] )
[[위상수학,topology]]
 
----
[[WpEn:Filter_(mathematics)]]
= Unix의 filter =
[[유닉스,Unix]]
에선 rel:
[[파이프,pipe]]
[[파이프라인,pipeline]]

----
rel
@@ -127,6 +204,7 @@
https://artofproblemsolving.com/wiki/index.php/Filter
[[WpEn:Filter_(mathematics)]]
[[WpKo:필터_(수학)]]
https://ncatlab.org/nlab/show/filter

분야는
[[집합론,set_theory]] [[위상수학,topology]](curr. [[위상,topology]] - [[VG:위상,topology]]) ...

거름망, 거르개, 거름종이, .. 암튼 거르는 도구인데

filter가 하는 행동이 필터링,filtering? { 신호,signal에 포함된 잡음이나 간섭 성분을 제거하거나, ... }
or 여과,filtration?

Sub: // 이건 추정,estimation으로 묶?
Bayes_filter WpEn:Bayes_filter
Kalman_filter WpEn:Kalman_filter

[edit]

1. CNN에서 filter

CNN,convolutional_neural_network
or
합성곱신경망,convolutional_neural_network,CNN
에서는 입력의 일부를 (보통 정사각형? - x, y 크기는 홀수인(그래서 중심이 어디인지 명확한)) 잘라내고 그걸 행렬,matrix과 elementwise product를 구하는 즉 아다마르_곱,Hadamard_product을 구하는 것을 반복하여 합성곱,convolution을 구하게 되는데... 이 행렬? chk
잘라낸 곳과 hadamard product 구하는 것은 convolution이 아니고 autocorrelation이고,
filter가 반대로 된 것(inverted?)과 구해야 convolution인데
CNN에서는 대부분 filter kernel이 대칭인 것을 쓰므로 그냥 구분없이 'convolutional' NN이라 한다는데....? chk

filter가 건너뛰는 step은 stride. - chk

Sub:
Haar-like_filter
image에서 features를 detect할 때 사용?
gaussian_filter
... Google:gaussian_filter
gaussian_derivative_filter (? del ok)
... Google:gaussian_derivative_filter
Laplacian of Gaussian (LoG) Filter
... Google:Laplacian of Gaussian Filter

... Google:CNN filter

TBW
kernel_filter Google:kernel_filter - syn?
receptive_filter Google:receptive_filter - receptive_field?

비교: (CNN의 다른 용어들)
channel - 채널,channel - cnn에서 image의 channel은 R, G, B를 뜻하는 듯? 항상?
stride - easy { 1 이상의 정수? }
padding - easy
kernel - 커널,kernel

[edit]

2. 신호, 신호처리에서 filter

[edit]

3. 제어에서 filter? (위와 merge?)

[edit]

4. Python filter

stmt?

파이썬,Python의 filter(함수, 반복가능한것)은
반복가능한것이 함수에 입력되었을 때 return_value 가 참,true인 것만 걸러내어 돌려준다? chk

[edit]

5. 수학(esp. order theory, lattice theory)의 filter

번역은?
KmsE:filter에선 그냥 필터로 함.
뜻(번역후보)는 여과 거름 .... 등등.

poset의 special subset, describing "large" or "eventual" elements. (we)

MKL
구조,structure
ultrafilter =,ultrafilter . ultrafilter
{
"ultra filter 극대필터, 초필터" via KmsE:ultra filter

주어진 poset $\displaystyle P$ 위의 ultrafilter는 P의 특정 subset인데,


극대필터,maximal_filter =극대필터,maximal_filter =,maximal_filter . 극대필터 maximal_filter
{
maximal filter
극대필터 via kms.

극대,maximal
maximal_element - 극대원 or 극대원소

"maximal filter 극대필터
maximal filterbase 극대필터기저" via KmsE:maximal filter
}// maximal filter ... Naver:maximal filter Ggl:maximal filter

minimal_filter ?

proper_filter =,proper_filter =,proper_filter . proper_filter
{
proper filter
"proper filter 진필터" via KmsE:proper filter
} // proper filter .... Naver:proper filter Ggl:proper filter Bing:proper filter

prime_filter =,prime_filter =,prime_filter . prime_filter
{
prime filter
prime filter
Sub:
completely_prime_filter
} // prime filter .... NN:prime filter Naver:prime filter Bing:prime filter

completely_prime_filter =,completely_prime_filter =,completely_prime_filter . completely_prime_filter
{
completely prime filter
https://ncatlab.org/nlab/show/completely prime filter
} // completely prime filter .... Naver:completely prime filter Bing:completely prime filter

MKL
Martin_measure =,Martin_measure . Martin_measure
{
Martin measure

https://en.wikipedia.org/wiki/Martin_measure
"In descriptive_set_theory =,descriptive_set_theory =,descriptive_set_theory . descriptive_set_theory { https://ko.wikipedia.org/wiki/기술적_집합론 https://en.wikipedia.org/wiki/Descriptive_set_theory 기술,description 집합론,set_theory } , the Martin measure is a filter(필터,filter) on the set of Turing_degrees of sets of natural numbers, named after Donald A. Martin{ https://en.wikipedia.org/wiki/Donald_A._Martin ...에 따른 업적은 주로 집합론,set_theory의 determinacy =,determinacy . determinacy
{ NdEn:determinacy WtEn:determinacy https://en.wikipedia.org/wiki/Determinacy 게임,game 게임이론,game_theory 전략,strategy perfect_information ... Google:determinacy } }. Under the axiom of determinacy(determinacy_axiom https://en.wikipedia.org/wiki/Axiom_of_determinacy ) it can be shown to be an ultrafilter."

(via Turing_degree is_a 디그리,degree)


comparison of filters
같은 집합에 대한 filter들간의 비교. 어떤 filter가 다른 filter보다
  • finer
  • coarser
  • strictly finer
  • strictly coarser
  • comparable
에 대해: https://planetmath.org/comparisonoffilters
Up: 비교,comparison

REL
filtration =,filtration =,filtration . filtration (of mathematics)
{
filtration

Not in KmsE:filtration as of 2024-07-05 ....

}// filtration .... NN:filtration


[edit]

6. Unix의 filter

필터,filter Modified on 2:21 am, October 27, 2024.
🎲 Random