August 13, 2020

Breast Cancer Classifier Project

1. It is common practice to split function parameters on separate lines when there are too many for a single line:

   1. Example: func(

      1. param1,

      2. param2,

      3. param3,

      4. param4

   2. )

2. Example: Store the results from a Python function returning arrays

   1. array = [1, 2, 3]

   2. first, second, third = array

3. training_data, validation_data, training_labels, validation_labels = train_test_split(breast_cancer_data.data, breast_cancer_data.target, test_size = 0.2, random_state = 100 )

4. Use the classifier with k = 3 and check the score:

   1. classifier.fit(training_data, training_lables)

   2. print(classifier.score(validation_data, validation_labels))

5. K was hardcoded to 3. What if we loop from 1 to 100 to see what K is most accurate?

   1. accuracies = [ ]

   2. for k in range(1, 101)

      1. classifier = KNeighborsClassifier(n_neighbors = k)

      2. classifier.fit(training_points, training_labels)

      3. accuracies.append(classifier.score(validation_data, validation_labels)

      4. print(accuracies)

   3. plt.plot(k_list, accuracies)

   4. plt.show( )

6. Overfitting - rely too much on training data - K is low

7. Underfitting - failure to learn from training data - K is too high