Getting better at digits

The Setting

Let us put together the code we had used in the class for image classification using digits.

import matplotlib.pyplot as plt
from sklearn import datasets, svm, metrics
from sklearn.linear_model import LogisticRegression

digits = datasets.load_digits()
n_samples = len(digits.images)
data = digits.images.reshape((n_samples, -1))

# Create a classifier
classifier = LogisticRegression()

# We learn the digits on the first half of the digits
classifier.fit(data[:n_samples // 2], digits.target[:n_samples // 2])

# Now predict the value of the digit on the second half:
expected = digits.target[n_samples // 2:]
predicted = classifier.predict(data[n_samples // 2:])

print("Accuracy: %s\n" % (metrics.accuracy_score(expected, predicted)))

The accuracy using Logistic Regression is 91.7%.

Problem Statement

Write a function called solution() which

• accepts the parameters X_train_digits and y_train_digits datasets (available in your environment) for training, and
• Use SVC algorithm
• returns the trained model as output

Hint

• We'll create X_train_digits, y_train_digits, X_test and y_test
• X_train_digits and y_train_digits will be available in student's environment
• X_test and y_test will be available to us to evaluate the output from student's function