Building a SciKit-Learn Pipeline

sklearn-pipeline.py

# Make ML-Pipeline Runs Reproducible
random_state=7

# Testing our Pipeline with an ML-Ready Dataset
from sklearn.datasets import load_iris
iris = load_iris()

# Assign ML-Input and Ml-Target
Xi = iris.data   # Input
yi = iris.target # Output

# Split Dataset into train/test for input/output
Xi_train, Xi_test, yi_train, yi_test = train_test_split(Xi, yi, random_state)

# loading ml libraries
from sklearn.neighbors import KNeighborsClassifier
from sklearn.tree import DecisionTreeClassifier
from sklearn.ensemble import RandomForestClassifier

# loading visualisation libraries
%matplotlib inline
import matplotlib.pyplot as plt
import scikitplot.plotters as skplt

# instantiate learning models
knn = KNeighborsClassifier(n_neighbors=2, random_state)
tree = DecisionTreeClassifier(max_leaf_nodes=3, random_state)
forest = RandomForestClassifier(n_estimators = 500, n_jobs = -1, random_state)

# train the models
knn.fit(Xi_train, yi_train)
tree.fit(Xi_train, yi_train)
forest.fit(Xi_train, yi_train)

# create prediction probabilities
pred1 = knn.predict(Xi_test)
pred1 = tree.predict(Xi_test)
pred1 = forest.predict(Xi_test)

# visualize learning curve
skplt.plot_learning_curve(knn, Xi_test, yi_test)
skplt.plot_learning_curve(tree, Xi_test, yi_test)
skplt.plot_learning_curve(forest, Xi_test, yi_test)