Enum UnivariateScoringFunction

Namespace

AiDotNet.Enums

Assembly

AiDotNet.dll

Defines the scoring functions available for univariate feature selection.

public enum UnivariateScoringFunction

Fields

ChiSquared = 0

Uses the Chi-Squared test to measure the dependence between categorical features and the target.

For Beginners: The Chi-Squared test compares the observed frequencies in your data with the frequencies you would expect if there was no relationship between the feature and target.

Higher scores indicate stronger relationships. This test works best with categorical data (data that falls into distinct categories like "red", "green", "blue").

FValue = 1

Uses ANOVA F-value to measure how well continuous features can distinguish between different classes.

For Beginners: ANOVA (Analysis of Variance) F-value measures whether the means of different groups are significantly different from each other.

For example, if you're predicting house prices in different neighborhoods, ANOVA can tell you if features like square footage vary significantly across neighborhoods. Higher F-values indicate that the feature is more useful for distinguishing between classes.

MutualInformation = 2

Uses Mutual Information to measure how much knowing one feature tells you about the target.

For Beginners: Mutual Information measures how much information a feature provides about the target variable. It answers the question: "If I know the value of this feature, how much does that help me predict the target?"

Higher scores mean the feature provides more information. Mutual Information works well with both categorical and continuous features, making it very versatile.

Remarks

For Beginners: These are different statistical tests that measure how much a single feature is related to the target variable you're trying to predict.

Each scoring function is best suited for different types of data: - Chi-Squared: Best for categorical features and categorical targets - ANOVA F-Value: Best for continuous features and categorical targets - Mutual Information: Works well for both categorical and continuous features with any target type