Python Scikit-Learn Cheat Sheet for Machine Learning

Let’s create a basic example using scikit-learn library which will be used to

• Load the data  
• Divide the data into train and test,
• Train your data using the KNN Algorithm and,
• Predict the result

>>> import numpy as np
>>> X = np.random.random((10,5))
>>> y = np.array(['M','M','F','F','M','F','M','M','F','F','F'])
>>> X[X < 0.7] = 0

Once the data is loaded, your next task would be split your dataset into training data and testing data.

>>> from sklearn.model_selection import train_test_split
>>> X_train, X_test, y_train, y_test = train_test_split(X, y, random_state=0)

Supervised learning is a type of machine learning that enables the model to predict future outcomes after they are trained on labelled data.

Unsupervised learning is a type of machine learning that enables the model to predict future outcomes without being trained on the labelled data.

# Linear Regression Algorithm
>>> from sklearn.linear_model import LinearRegression
>>> lr = LinearRegression(normalize=True)
# Naive Bayes Algorithm
>>> from sklearn.naive_bayes import GaussianNB
>>> gnb = GaussianNB()
# KNN Algorithm
>>> from sklearn import neighbors
>>> knn = neighbors.KNeighborsClassifier(n_neighbors=5)
# Support Vector Machines (SVM)
>>> from sklearn.svm import SVC
>>> svc = SVC(kernel='linear’)

# Principal Component Analysis (PCA)
>>> from sklearn.decomposition import PCA
>>> pca = PCA(n_components=0.95)
# K Means Clustering Algorithm
>>> from sklearn.cluster import KMeans
>>> k_means = KMeans(n_clusters=3, random_state=0)

Fitting is a measure of how well a machine learning model generalizes to similar data to that on which it was trained

# For Supervised learning
>>> lr.fit(X, y) #Fits data into the model
>>> knn.fit(X_train, y_train)
>>> svc.fit(X_train, y_train)
# For Unsupervised Learning
>>> k_means.fit(X_train)#Fits data into the model
>>> pca_model = pca.fit_transform(X_train) #Fit to data, then transform it

Fitting is a measure of how well a machine learning model generalizes to similar data to that on which it was trained

# For Supervised learning
>>> y_pred=svc.predict(np.random((2,5))) #predict label
>>> y_pred=lr.predict(x_test) #predict label
>>> y_pred=knn.predict_proba(x_test)#estimate probablity of a label
# For Unsupervised Learning
>>> y_pred=k_means.predict(x_test) #predict labels in clustering algorithm

The sklearn.metrics module implements several loss, score, and utility functions to measure classification performance. 

# Mean Absolute Error
>>> from sklearn.metrics import mean_absolute_error
>>> y_true = [3, -0.5, 2]
>>> mean_absolute_error(y_true, y_pred)
# Mean Squared Error
>>> from sklearn.metrics import mean_squared_error
>>> mean_squared_error(y_test, y_pred)
# R² Score
>>> from sklearn.metrics import r2_score
>>> r2_score(y_true, y_pred)

The sklearn.metrics module implements several loss, score, and utility functions to measure regression performance. 

# Accuracy Score
>>> knn.score(X_test, y_test)
>>> from sklearn.metrics import accuracy_score
>>> accuracy_score(y_test, y_pred)
# Classification Report
>>> from sklearn.metrics import classification_report
>>> print(classification_report(y_test, y_pred))
# Confusion Matrix
>>> from sklearn.metrics import confusion_matrix
>>> print(confusion_matrix(y_test, y_pred))

# Adjusted Rand Index
>>> from sklearn.metrics import adjusted_rand_score
>>> adjusted_rand_score(y_true, y_pred)
# Homogeneity
>>> from sklearn.metrics import homogeneity_score
>>> homogeneity_score(y_true, y_pred)
# V-measure
>>> from sklearn.metrics import v_measure_score
>>> metrics.v_measure_score(y_true, y_pred)

>>> from sklearn.cross_validation import cross_val_score 
>>> print(cross_val_score(knn, X_train, y_train, cv=4)) 
>>> print(cross_val_score(lr, X, y, cv=2))

GridSearchCV implements a “fit” and a “score” method. It also implements “predict”, “predict_proba”, “decision_function”, “transform” and “inverse_transform” if they are implemented in the estimator used.

>>> from sklearn.grid_search import GridSearchCV
>>> params = {"n_neighbors": np.arange(1,3), "metric": ["euclidean", "cityblock"]}
>>> grid = GridSearchCV(estimator=knn, param_grid=params)
>>> grid.fit(X_train, y_train)
>>> print(grid.best_score_)
>>> print(grid.best_estimator_.n_neighbors)

RandomizedSearchCV performs the random search on hyper parameters. In contrast to GridSearchCV, not all parameter values are tried out, but rather a fixed number of parameter settings is sampled from the specified distributions. The number of parameter settings that are tried is given by n_iter.

>>> from sklearn.grid_search import RandomizedSearchCV
>>> params = {"n_neighbours": range(1,5), "weights":["uniform", "distance"]}
>>>rserach = RandomizedSearchCV(estimator=knn,param_distribution=params, cv=4, n_iter=8, random_state=5)
>>> rsearch.fit(X_train, Y_train)
>>> print(rsearch.best_score)