Stacking in ML

Stacking Architecture

#!/usr/bin/env python

# coding: utf-8

# In[38]:

import pandas as pd

import numpy as np

from sklearn.model_selection import train_test_split

from sklearn.model_selection import GridSearchCV

from sklearn import tree

from sklearn.ensemble import RandomForestClassifier

from sklearn.neighbors import KNeighborsClassifier

from sklearn.svm import SVC

# In[8]:

df = pd.read_csv('diabetes.csv')

# In[10]:

X = df.drop(columns = 'Outcome')

y = df['Outcome']

# In[12]:

train, val_train, test, val_test = train_test_split(X,y, random_state = 355, test_size = 0.5 )

# In[13]:

X_train, X_test, y_train, y_test = train_test_split(train, test, test_size = 0.2, random_state = 355)

# In[15]:

knn = KNeighborsClassifier()

# In[17]:

knn.fit(X_train,y_train)

# In[18]:

knn.score(X_train, y_train)

# In[19]:

svm = SVC()

# In[20]:

svm.fit(X_train, y_train)

# In[21]:

svm.score(X_train, y_train)

# In[25]:

predict_val1 = knn.predict(val_train)

predict_val2 = svm.predict(val_train)

# In[27]:

predict_val = np.column_stack((predict_val1, predict_val2))

# In[28]:

predict_val

# In[33]:

predict_test1 = knn.predict(X_test)

predict_test2 = svm.predict(X_test)

# In[34]:

predict_test = np.column_stack((predict_test1, predict_test2))

# In[ ]:

# In[29]:

rand_clf = RandomForestClassifier()

# In[30]:

rand_clf.fit(predict_val, val_test)

# In[35]:

rand_clf.score(predict_test,y_test)

# In[40]:

grid_parm = {

    

    "n_estimators":[90, 100, 115],

    "criterion": ['gini', 'entropy'],

    'min_samples_leaf':[1,2,3,4,5],

    "min_samples_split":[4,5,6,7,8],

    "max_features" :['auto', 'log2']

}

# In[45]:

grid_search = GridSearchCV(estimator= rand_clf, param_grid = grid_parm, cv=5, n_jobs = -1, verbose=3)

# In[46]:

grid_search.fit(predict_val, val_test)

# In[47]:

grid_search.best_params_

# In[50]:

rand_clf = RandomForestClassifier(criterion = 'gini', max_features = 'auto', min_samples_leaf=1, min_samples_split=4, n_estimators=90)

# In[51]:

rand_clf.fit(predict_val, val_test)

# In[52]:

rand_clf.score(predict_test, y_test)

# In[ ]: