코딩하는 임초얀

Intrinsic dimension 본문

Wiki 한글 번역

Intrinsic dimension

초얀 2022. 9. 23. 10:00

The intrinsic dimension for a data set can be thought of as the number of variables needed in a minimal representation of the data. Similarly, in signal processing of multidimensional signals, the intrinsic dimension of the signal describes how many variables are needed to generate a good approximation of the signal.

 

데이터셋의 intrinsic dimension데이터의 최소 표현에 필요한 변수의 수로 생각할 수 있다. 마찬가지로, 다차원 신호의 신호 처리에서, 신호의 intrinsic dimension은 신호를 잘 근사하기 위해 얼마나 많은 변수가 필요한지를 말한다.

 


When estimating intrinsic dimension, however, a slightly broader definition based on manifold dimension is often used, where a representation in the intrinsic dimension does only need to exist locally. Such intrinsic dimension estimation methods can thus handle data sets with different intrinsic dimensions in different parts of the data set. This is often referred to as local intrinsic dimensionality.


그러나 intrinsic dimension을 추정할 때, 다양체 차원(manifold dimension)에 기초한 약간 더 넓은 정의가 종종 사용되며, intrinsic dimension의 표현은 국소적으로만 존재하면 된다. 따라서 이러한 intrinsic dimension 추정 방법은 데이터셋의 여러 부분들이 서로 다른 intrinsic dimension을 가지는 상황을 처리할 수 있다. 이것을 local intrinsic dimensionality라고 부른다.

 


The intrinsic dimension can be used as a lower bound of what dimension it is possible to compress a data set into through dimension reduction, but it can also be used as a measure of the complexity of the data set or signal. For a data set or signal of N variables, its intrinsic dimension M satisfies 0 ≤ M ≤ N, although estimators may yield higher values.


Intrinsic dimension차원 축소(dimension reduction)를 통해 데이터셋을 압축할 수 있는 차원의 최솟값으로 사용될 수도 있지만, 데이터셋 또는 신호의 복잡도를 측정하는 척도로 사용될 수도 있다. 변수가 N개인 데이터셋 또는 신호의 경우, (추정기가 더 높은 값을 산출할 수도 있지만) intrinsic dimension M은 0 ≤ M ≤ N을 만족한다.

'Wiki 한글 번역' 카테고리의 다른 글

Bellman Equation  (0) 2023.03.03
Radiance  (0) 2022.09.28
Comments