2022 Machine Learning PHW1

Various algorithm test automation functions for one dataset

Result Example

Document

Class getBest5(dataset= 'dataset path')

When a user inserts a dataset, the dataset is preprocessed and the best combination of hyper parameters, scaler, and model is found and Top5 is output. There are 3 scalers [MinMaxScaler(), StandardScaler(), RobustScaler()] and 4 algorithms [Logistic Regression(), Entropy Classifier(), Gini Classifier(), SVC()] and many hyper parameters for each algorithm Included.

LogisticRegression

C=[0.001, 0.01, 0.1, 1, 10, 25, 50, 100]

max_iter=[50,100,150]

lver=['newton-cg', 'lbfgs', 'liblinear']

DecisionTreeClassifier

criterion = ['entropy']

max_depth = [2,4,6,8,10,12]

min_samples_leaf = [1, 2, 5, 10]

min_samples_split = [2, 5, 10, 15]

max_features = ['auto', 'sqrt','log2']

Gini Classifier

criterion = ['gini']

max_depth = [2,4,6,8,10,12]

min_samples_leaf = [1, 2, 5, 10]

min_samples_split = [2, 5, 10, 15]

max_features = ['auto', 'sqrt','log2']

SVC

C=[0.001, 0.01, 0.1, 1, 10, 25, 50, 100]

kernel=['linear', 'sigmoid', 'rbf', 'poly']

gamma=[0.001, 0.01, 0.1, 1, 10, 25, 50, 100]

Examples

dataset = 'breastCancer.csv'

print(getBest5(dataset))

Methods

preprocessing(dataset = 'dataset path'):

parameter:: dataset

The path of the dataset you want to preprocess

return:: X_train, X_test, y_train, y_test

Preprocessing the dataset and dividing it into a train set and a test set

modeling(X_train, X_test, y_train, y_test):

parameter:: X_train, X_test, y_train, y_test

The train set and test set of the dataset you want to experiment with

Return:: final_top5.values.tolist()

Combination of scaler, algorithm, and hyper parameters of the final Top 5 composed of a list