What Is The Difference Between Dimensions And Features In Machine Learning

In car learning, the word "dimension" should exist a loftier-frequency discussion, which often appears in people's vision. For instance, random woods is built by random feature extraction to avoid high-dimensional calculation; for example, sklearns must be at least 2-dimensional when introducing feature matrix; the purpose of feature selection is through descent. Dimension reduces the computational cost of the algorithm… These languages were normally used past me until one mean solar day a little friend asked me, "What is dimension?" I…

Afterward careful consideration, I sum upwards as follows:

i. For arrays and Series

nearArrays and SeriesFor instance,Dimensions are the results of functional shapes, and a few numbers are returned in shapes, that is, the dimensions.。 Information other than indexing is called 1-dimensional (when shape returns the number of data on the unique dimension), and 2-dimensional (shape returns row x) is also called table. A table is at most 2-dimensional, and a complex table constitutes a higher dimension. When there are ii tables with three rows and iv columns in an array, shape returns (higher dimensions, rows, columns). When at that place are two groups of tables with three rows and four columns in the array, the data is four-dimensional and the shape returns (2,ii,3,four).

Each table in an array tin be oneCharacteristic MatrixOr a DataFrame, where these structures always accept only one table, so there must be rows, where rows are samples and columns are features. For each tabular array,Dimensions refer to the number of samples or features, which are not specifically specified, but the number of features.。 In improver to index, one characteristic is i-dimensional, two features are two-dimensional, and North features are n-dimensional.

two. For images

Dimensions are the number of characteristic vectors in an image.Eigenvector can be understood as coordinate centrality, a feature vector defines a straight line, is one-dimensional, two mutually perpendicular feature vectors define a aeroplane, that is, a rectangular coordinate system, that is, 2-dimensional, three mutually perpendicular feature vectors define a space, that is, a three-dimensional rectangular coordinate organization, that is, three-dimensional. More than three feature vectors are perpendicular to each other, which defines a loftier-dimensional space that the homo eye cannot see or imagine.

iii. Dimension Reduction in Dimension Reduction Algorithms

In dimension reduction algorithm, "dimension reduction" refers to the reduction of the number of features in the feature matrix.As we said in final week's lecture, the purpose of dimensionality reduction is to make it possible for us to reduce dimensionality.The algorithm has faster operation and better result.Only in that location is another demand:Data visualization。 From the graph above, we tin can run across that the dimensions of paradigm and feature matrix tin can correspond to each other, that is, a characteristic corresponds to a characteristic vector and a coordinate centrality. Therefore, the three-dimensional and below feature matrix can be visualized, which can help us quickly understand the distribution of data, while the iii-dimensional and above feature matrix can non exist visualized, and the nature of data is difficult to sympathise.

Well, that'south the summary of dimensionality reduction. If you accept new ideas, you are welcome to discuss them together.~