Table of Contents

Interface IFitnessCalculator<T, TInput, TOutput>

Namespace
AiDotNet.Interfaces
Assembly
AiDotNet.dll

Defines an interface for calculating how well a machine learning model performs.

public interface IFitnessCalculator<T, TInput, TOutput>

Type Parameters

T

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

TInput
TOutput

Remarks

For Beginners: This interface helps measure how "fit" or effective your machine learning model is.

In machine learning, we need a way to measure how good our model is at making predictions. This is similar to how we might grade a test - we need a scoring system.

Different types of problems need different scoring methods:

  • For predicting house prices, we might measure how close our predictions are to actual prices
  • For classifying emails as spam/not spam, we might count how many emails we classified correctly

The "fitness score" is this measurement of how well the model performs. Some important points:

  • Sometimes higher scores are better (like accuracy: 95% is better than 90%)
  • Sometimes lower scores are better (like error: 5% error is better than 10% error)
  • The score helps us compare different models to choose the best one

This interface provides methods to calculate these fitness scores in a standardized way.

Properties

IsHigherScoreBetter

Indicates whether higher fitness scores represent better performance.

bool IsHigherScoreBetter { get; }

Property Value

bool

Remarks

For Beginners: This property tells you whether bigger numbers mean better performance.

Different metrics work in opposite ways:

  • For some metrics like accuracy, higher is better (95% accuracy is better than 90%)
  • For other metrics like error rate, lower is better (5% error is better than 10%)

This property returns:

  • true: if higher scores mean better performance (like accuracy)
  • false: if lower scores mean better performance (like error)

Knowing this helps you correctly interpret and compare fitness scores.

Methods

CalculateFitnessScore(DataSetStats<T, TInput, TOutput>)

Calculates a fitness score based on basic dataset statistics.

T CalculateFitnessScore(DataSetStats<T, TInput, TOutput> dataSet)

Parameters

dataSet DataSetStats<T, TInput, TOutput>

Statistical information about model performance on a dataset.

Returns

T

A fitness score representing how well the model performs.

Remarks

For Beginners: This method grades your model based on simplified performance statistics.

The input parameter:

  • dataSet: Contains basic statistics about how your model performed This is a simpler version of the evaluation data used in the other method

This method is useful when you only have access to summary statistics rather than detailed evaluation data. It still produces a fitness score that measures how good your model is, just using less detailed information.

For example, if you're building a spam filter, this might calculate accuracy based on how many emails were correctly classified as spam or not spam.

CalculateFitnessScore(ModelEvaluationData<T, TInput, TOutput>)

Calculates a fitness score based on comprehensive model evaluation data.

T CalculateFitnessScore(ModelEvaluationData<T, TInput, TOutput> evaluationData)

Parameters

evaluationData ModelEvaluationData<T, TInput, TOutput>

Data containing model performance metrics on training and validation datasets.

Returns

T

A fitness score representing how well the model performs.

Remarks

For Beginners: This method grades your model based on detailed performance information.

The input parameter:

  • evaluationData: Contains various measurements about how your model performed on different data This typically includes information from both:
    • Training data (data your model learned from)
    • Validation data (new data your model hasn't seen before)

The method processes these measurements and returns a single number (the fitness score) that represents how good your model is. This makes it easy to compare different models or different versions of the same model.

For example, if you're predicting house prices, this might calculate the average percentage error in your predictions compared to actual prices.

IsBetterFitness(T, T)

Compares two fitness scores and determines if the current score is better than the best score so far.

bool IsBetterFitness(T currentFitness, T bestFitness)

Parameters

currentFitness T

The fitness score to evaluate.

bestFitness T

The best fitness score found so far.

Returns

bool

True if the current fitness score is better than the best fitness score; otherwise, false.

Remarks

For Beginners: This method compares two scores and tells you if the new one is better.

When comparing models, you need to know which one performs better. This method handles the comparison logic for you, taking into account whether higher or lower scores are better.

The input parameters:

  • currentFitness: The new score you want to evaluate
  • bestFitness: The best score you've seen so far

The method returns:

  • true: if the current fitness is better than the best fitness
  • false: if the current fitness is worse than or equal to the best fitness

This is particularly useful in algorithms that try many different models and need to keep track of which one is best.

Edit this page