MlQuery

Functions

• UserFilePath() : Takes the file path from the user
• selectFeatures() : Takes the features and the target column from the user

Preprocessing

• replaceNUL() : This function either drops the complete column if more than 50% of the data is missing or it would replace it withe either 0, mean, mode or median.
• featGet(data, features, target) : returns values of X and y according to the choosen features and target columns.
• oneHotEncoding(X, columnNumber) : This function performs one Hot encoding, given columnNumebr.
• split_dataset(X, y, testSize) : This function return X_train, X_test, y_train, y_test according to the given testSize
• standard(X_train, X_test) : This function Standardizes X_train, X_test and return them.
• normalize(X_train, X_test) : This function Normalizes X_train, X_test and returns them.

Training Models

• logReg(X_train, X_test, y_train) : performs logistice regression and return y_pred.
• svcModel(X_train, X_test, y_train) : performs SVC and returns y_pred.
• knnModel(X_train, X_test, y_train, n) : performs knn, where n = n_neighbors and return y_pred.

Accuracy score

• accuracy(y_pred, y_test) : performs accuracy prediction and returns the accuracy score.

Sample Input

Enter location of the file: E:/ML PROJECTS/Arya's SInking titanic/SinkingTitanic/train.csv Takes the file path

Enter the names of features : Pclass,Sex,Age Takes the features for X

Enter the target between '' : Survived Takes the column for y

How do u want to replace the null values for feature: Age Finds the column withe null value and ask if you want to perform mean, median or mode with missing data

1. with 0

2.with mean

3.with median

4.with mode

2 chose option 2

Do you want to perform One Hot Encoding?[y/n] option to perform One Hot Encoding n

Do you want to perform Label Encoding?[y/n] Option to perform Label Encoding n

Do you want to perform Standardization ?[y/n]y Option to perform Standardization

Do you want to perform Normalization ?[y/n]n Option to perform Normalization

option to choose from regression or classification

Enter 1. for Regression:

Enter 2. for Classification:

2

Enter the value for K7

            Logistic Regression  K-nearest neighbours

Accuracy              0.12037              0.015856
      
      Logistic Regression  K-nearest neighbours

Accuracy             0.088649              0.109043
      
      Logistic Regression  K-nearest neighbours

Accuracy             0.169979              0.265189
      
      Logistic Regression  K-nearest neighbours

Accuracy             0.112783               0.22028
      
      Logistic Regression  K-nearest neighbours

Accuracy             0.151037              0.029713
      
      Logistic Regression  K-nearest neighbours

Accuracy             0.060074             -0.051049
      
      Logistic Regression  K-nearest neighbours

Accuracy             0.060917              0.107806
      
      Logistic Regression  K-nearest neighbours

Accuracy              0.03062              0.040513