Interface IOutlierRemoval<T, TInput, TOutput>

Namespace

AiDotNet.Interfaces

Assembly

AiDotNet.dll

Defines methods for detecting and removing outliers from datasets.

public interface IOutlierRemoval<T, TInput, TOutput>

Type Parameters

T

The numeric type used for calculations (e.g., float, double).

TInput

TOutput

Remarks

For Beginners: Outliers are unusual data points that differ significantly from most of your data. These unusual values can negatively impact machine learning models by skewing results.

This interface provides a standard way to implement different outlier detection and removal techniques. By removing outliers, you can often improve the accuracy and reliability of your machine learning models.

Methods

RemoveOutliers(TInput, TOutput)

Removes outliers from the input data and returns the cleaned dataset.

(TInput CleanedInputs, TOutput CleanedOutputs) RemoveOutliers(TInput inputs, TOutput outputs)

Parameters

inputs TInput

The input feature matrix where each row represents a data point and each column represents a feature.

outputs TOutput

The target values vector where each element corresponds to a row in the input matrix.

Returns

(TInput Input, TOutput Output)

A tuple containing:

• CleanedInputs: A matrix of input features with outliers removed
• CleanedOutputs: A vector of output values corresponding to the cleaned inputs

Remarks

For Beginners: This method analyzes your dataset to find unusual values (outliers) and removes them.

The inputs parameter is a matrix where:

• Each row represents one sample or data point
• Each column represents one feature or attribute

The outputs parameter is a vector where:

• Each element is the target value or label for the corresponding row in the inputs matrix

After processing, the method returns both the cleaned inputs and outputs with outliers removed, maintaining the relationship between input features and their corresponding output values.