

In [ ]:

    
print('ex01')



In [ ]:

    
import numpy as np
import pandas as pd

3.1 Problem description

Try to build a classifier for the MNIST dataset that achieves over 97% accuracy on the test set. Hint: the KNeighborsClassifier works quite well for this task; you just need to find good hyperparameter values (try a grid search on the weights and n_neighbors hyperparameters).

Load the data



In [ ]:

    
from scipy.io import loadmat
mnist = loadmat('./datasets/mnist-original.mat')



In [ ]:

    
mnist



In [ ]:

    
X, y = mnist['data'], mnist['label']



In [ ]:

    
X = X.T
X.shape



In [ ]:

    
y = y.T
y.shape



In [ ]:

    
type(y)



In [ ]:

    
%matplotlib inline
import matplotlib
import matplotlib.pyplot as plt

Split test and training data



In [ ]:

    
X_train, X_test, y_train, y_test = X[:60000], X[60000:], y[:60000], y[60000:]



In [ ]:

    
len(X_train)



In [ ]:

    
shuffle_index = np.random.permutation(len(X_train))
X_train, y_train = X_train[shuffle_index], y_train[shuffle_index]

3.2 Training a Random Forest Classifier for baseline

The reason to use Random Forest Classifier is it runs faster than linear model



In [ ]:

    
from sklearn.ensemble import RandomForestClassifier

forest_clf = RandomForestClassifier(random_state=42)



In [ ]:

    
forest_clf.fit(X_train, y_train)



In [ ]:

    
forest_pred = forest_clf.predict(X_test)



In [ ]:

    
forest_pred = forest_pred.reshape(10000,1)



In [ ]:

    
accuracy = (forest_pred == y_test).sum() / len(y_test)
print(accuracy)

3.3 Training a `KNeighborsClassifier` Classifier with default settings

Seems like we have to have n_jobs = 1 so the prediction runs within reasonable time.



In [ ]:

    
from sklearn.neighbors import KNeighborsClassifier



In [ ]:

    
knn_clf = KNeighborsClassifier(n_jobs=-1)



In [ ]:

    
knn_clf.fit(X_train, y_train)



In [ ]:

    
knn_clf.predict([X_test[0]])



In [ ]:

    
# for i in range(1000):
#     knn_clf.predict([X_test[i]])



In [ ]:

    
knn_pred = knn_clf.predict(X_test)



In [ ]:

    
knn_pred = knn_pred.reshape(10000, 1)
accuracy = (knn_pred == y_test).sum() / len(y_test)
print(accuracy)

3.4 `GridSearchCV`



In [ ]:

    
from sklearn.model_selection import GridSearchCV



In [ ]:

    
param_grid = [
    {'n_jobs': [-1], 'n_neighbors': [3, 5, 11, 19], 'weights': ['uniform', 'distance']}
]



In [ ]:

    
grid_search = GridSearchCV(knn_clf, param_grid, cv=3, scoring='accuracy', n_jobs=-1)



In [ ]:

    
grid_search.fit(X_train, y_train)