prevhlib 0.0.9

Data Mining package implementing the PrevhClassifier algorithm.

Prevh

This package implements the Prevh classification algorithm.

The algorithm is based in the follow research Pages 71-76.

Package Documentation.

User Guide

This package can be installed with the following command: pip install prevhlib

Dataset

The file must be a CSV file and the header must be included.

The columns must be in the following order:

• The features columns;
• The label column;
• The relevance column (Optional).

feature1,feature2,feature3,label,relevance
10,10,10,Blue,1.0
15,15,15,Blue,1.0
20,20,20,Blue,1.0
45,45,45,Green,1.0
50,50,50,Green,1.0
55,55,55,Green,1.0
80,80,80,Red,1.0
85,85,85,Red,1.0
90,90,90,Red,1.0

Python example:

import prevh as ph
import pandas as pd
# Creates the classifier
prevhClass = PrevhClassifier(pd.read_csv("irisDataCSV.csv",","))
# Label recurrence in the dataset (Important to use KNR method)
print(prevhClass.labelCount)
# Rows count in the dataset (Important to use KNN method)
print(prevhClass.rowsCount)
# Calculate the dataset score using the TrainTestSplit and KFold Cross-Validation methods
TrainTestSplitScore = prevhclass.calculateScore("TrainTestSplit", algorithm="KNN", k=4, train_size=0.8, seed=42)
KfoldScore = prevhclass.calculateScore("KFold", algorithm="KNR", k=35, n_splits=15, seed=42)
print("TrainTestSplitScore:", TrainTestSplitScore)
print("KFoldScore:", KfoldScore)

Change Log

0.0.9 (26/03/2022)

MAJOR CHANGES

• Added simple predict funtionality
• Added KNR and KNN space delimitation methods
• Added TestTrainSplit and KFold-CrossValidation score methods to evalueate databases