Machine Learning Engineer Nanodegree

Supervised Learning

Project: Building a Student Intervention System

Welcome to the second project of the Machine Learning Engineer Nanodegree! In this notebook, some template code has already been provided for you, and it will be your job to implement the additional functionality necessary to successfully complete this project. Sections that begin with 'Implementation' in the header indicate that the following block of code will require additional functionality which you must provide. Instructions will be provided for each section and the specifics of the implementation are marked in the code block with a 'TODO' statement. Please be sure to read the instructions carefully!

In addition to implementing code, there will be questions that you must answer which relate to the project and your implementation. Each section where you will answer a question is preceded by a 'Question X' header. Carefully read each question and provide thorough answers in the following text boxes that begin with 'Answer:'. Your project submission will be evaluated based on your answers to each of the questions and the implementation you provide.

Note: Code and Markdown cells can be executed using the Shift + Enter keyboard shortcut. In addition, Markdown cells can be edited by typically double-clicking the cell to enter edit mode.

Question 1 - Classification vs. Regression

Your goal for this project is to identify students who might need early intervention before they fail to graduate. Which type of supervised learning problem is this, classification or regression? Why?

Answer: Identifying students who require intervention is a classification problem. This is because we are predicting whether or not a student requires intervention or not which is a discrete value. A regression problem is one that has continuous values such as predicting house value.

Exploring the Data

Run the code cell below to load necessary Python libraries and load the student data. Note that the last column from this dataset, 'passed', will be our target label (whether the student graduated or didn't graduate). All other columns are features about each student.


In [5]:
# Import libraries
import numpy as np
import pandas as pd
from time import time
from sklearn.metrics import f1_score

# Read student data
student_data = pd.read_csv("student-data.csv")
print "Student data read successfully!"


Student data read successfully!

Implementation: Data Exploration

Let's begin by investigating the dataset to determine how many students we have information on, and learn about the graduation rate among these students. In the code cell below, you will need to compute the following:

  • The total number of students, n_students.
  • The total number of features for each student, n_features.
  • The number of those students who passed, n_passed.
  • The number of those students who failed, n_failed.
  • The graduation rate of the class, grad_rate, in percent (%).

In [6]:
# TODO: Calculate number of students
n_students = len(student_data)

# TODO: Calculate number of features
n_features = len(student_data.columns)

# TODO: Calculate passing studevalue_counts()['yes']

n_passed = student_data['passed'].value_counts()['yes']

# TODO: Calculate failing students
n_failed = student_data['passed'].value_counts()['no']

# TODO: Calculate graduation rate
grad_rate = (np.true_divide(n_passed,n_students))*100


# Print the results
print "Total number of students: {}".format(n_students)
print "Number of features: {}".format(n_features)
print "Number of students who passed: {}".format(n_passed)
print "Number of students who failed: {}".format(n_failed)
print "Graduation rate of the class: {:.2f}%".format(grad_rate)


Total number of students: 395
Number of features: 31
Number of students who passed: 265
Number of students who failed: 130
Graduation rate of the class: 67.09%

Preparing the Data

In this section, we will prepare the data for modeling, training and testing.

Identify feature and target columns

It is often the case that the data you obtain contains non-numeric features. This can be a problem, as most machine learning algorithms expect numeric data to perform computations with.

Run the code cell below to separate the student data into feature and target columns to see if any features are non-numeric.


In [7]:
# Extract feature columns
feature_cols = list(student_data.columns[:-1])

# Extract target column 'passed'
target_col = student_data.columns[-1] 

# Show the list of columns
print "Feature columns:\n{}".format(feature_cols)
print "\nTarget column: {}".format(target_col)

# Separate the data into feature data and target data (X_all and y_all, respectively)
X_all = student_data[feature_cols]
y_all = student_data[target_col]

# Show the feature information by printing the first five rows
print "\nFeature values:"
print X_all.head()


Feature columns:
['school', 'sex', 'age', 'address', 'famsize', 'Pstatus', 'Medu', 'Fedu', 'Mjob', 'Fjob', 'reason', 'guardian', 'traveltime', 'studytime', 'failures', 'schoolsup', 'famsup', 'paid', 'activities', 'nursery', 'higher', 'internet', 'romantic', 'famrel', 'freetime', 'goout', 'Dalc', 'Walc', 'health', 'absences']

Target column: passed

Feature values:
  school sex  age address famsize Pstatus  Medu  Fedu     Mjob      Fjob  \
0     GP   F   18       U     GT3       A     4     4  at_home   teacher   
1     GP   F   17       U     GT3       T     1     1  at_home     other   
2     GP   F   15       U     LE3       T     1     1  at_home     other   
3     GP   F   15       U     GT3       T     4     2   health  services   
4     GP   F   16       U     GT3       T     3     3    other     other   

    ...    higher internet  romantic  famrel  freetime goout Dalc Walc health  \
0   ...       yes       no        no       4         3     4    1    1      3   
1   ...       yes      yes        no       5         3     3    1    1      3   
2   ...       yes      yes        no       4         3     2    2    3      3   
3   ...       yes      yes       yes       3         2     2    1    1      5   
4   ...       yes       no        no       4         3     2    1    2      5   

  absences  
0        6  
1        4  
2       10  
3        2  
4        4  

[5 rows x 30 columns]

Preprocess Feature Columns

As you can see, there are several non-numeric columns that need to be converted! Many of them are simply yes/no, e.g. internet. These can be reasonably converted into 1/0 (binary) values.

Other columns, like Mjob and Fjob, have more than two values, and are known as categorical variables. The recommended way to handle such a column is to create as many columns as possible values (e.g. Fjob_teacher, Fjob_other, Fjob_services, etc.), and assign a 1 to one of them and 0 to all others.

These generated columns are sometimes called dummy variables, and we will use the pandas.get_dummies() function to perform this transformation. Run the code cell below to perform the preprocessing routine discussed in this section.


In [8]:
def preprocess_features(X):
    ''' Preprocesses the student data and converts non-numeric binary variables into
        binary (0/1) variables. Converts categorical variables into dummy variables. '''
    
    # Initialize new output DataFrame
    output = pd.DataFrame(index = X.index)

    # Investigate each feature column for the data
    for col, col_data in X.iteritems():
        
        # If data type is non-numeric, replace all yes/no values with 1/0
        if col_data.dtype == object:
            col_data = col_data.replace(['yes', 'no'], [1, 0])

        # If data type is categorical, convert to dummy variables
        if col_data.dtype == object:
            # Example: 'school' => 'school_GP' and 'school_MS'
            col_data = pd.get_dummies(col_data, prefix = col)  
        
        # Collect the revised columns
        output = output.join(col_data)
    
    return output

X_all = preprocess_features(X_all)
print "Processed feature columns ({} total features):\n{}".format(len(X_all.columns), list(X_all.columns))


Processed feature columns (48 total features):
['school_GP', 'school_MS', 'sex_F', 'sex_M', 'age', 'address_R', 'address_U', 'famsize_GT3', 'famsize_LE3', 'Pstatus_A', 'Pstatus_T', 'Medu', 'Fedu', 'Mjob_at_home', 'Mjob_health', 'Mjob_other', 'Mjob_services', 'Mjob_teacher', 'Fjob_at_home', 'Fjob_health', 'Fjob_other', 'Fjob_services', 'Fjob_teacher', 'reason_course', 'reason_home', 'reason_other', 'reason_reputation', 'guardian_father', 'guardian_mother', 'guardian_other', 'traveltime', 'studytime', 'failures', 'schoolsup', 'famsup', 'paid', 'activities', 'nursery', 'higher', 'internet', 'romantic', 'famrel', 'freetime', 'goout', 'Dalc', 'Walc', 'health', 'absences']

Implementation: Training and Testing Data Split

So far, we have converted all categorical features into numeric values. For the next step, we split the data (both features and corresponding labels) into training and test sets. In the following code cell below, you will need to implement the following:

  • Randomly shuffle and split the data (X_all, y_all) into training and testing subsets.
    • Use 300 training points (approximately 75%) and 95 testing points (approximately 25%).
    • Set a random_state for the function(s) you use, if provided.
    • Store the results in X_train, X_test, y_train, and y_test.

In [9]:
# TODO: Import any additional functionality you may need here
from sklearn.model_selection import train_test_split

# TODO: Set the number of training points
num_train = 300

# Set the number of testing points
num_test = X_all.shape[0] - num_train

# TODO: Shuffle and split the dataset into the number of training and testing points above
X_train, X_test = train_test_split(X_all, train_size=num_train, test_size=num_test)
y_train, y_test = train_test_split(y_all, train_size=num_train, test_size=num_test)

# Show the results of the split
print "Training set has {} samples.".format(X_train.shape[0])
print "Testing set has {} samples.".format(X_test.shape[0])


Training set has 300 samples.
Testing set has 95 samples.

Training and Evaluating Models

In this section, you will choose 3 supervised learning models that are appropriate for this problem and available in scikit-learn. You will first discuss the reasoning behind choosing these three models by considering what you know about the data and each model's strengths and weaknesses. You will then fit the model to varying sizes of training data (100 data points, 200 data points, and 300 data points) and measure the F1 score. You will need to produce three tables (one for each model) that shows the training set size, training time, prediction time, F1 score on the training set, and F1 score on the testing set.

The following supervised learning models are currently available in scikit-learn that you may choose from:

  • Gaussian Naive Bayes (GaussianNB)
  • Decision Trees
  • Ensemble Methods (Bagging, AdaBoost, Random Forest, Gradient Boosting)
  • K-Nearest Neighbors (KNeighbors)
  • Stochastic Gradient Descent (SGDC)
  • Support Vector Machines (SVM)
  • Logistic Regression

Question 2 - Model Application

List three supervised learning models that are appropriate for this problem. For each model chosen

  • Describe one real-world application in industry where the model can be applied. (You may need to do a small bit of research for this — give references!)
  • What are the strengths of the model; when does it perform well?
  • What are the weaknesses of the model; when does it perform poorly?
  • What makes this model a good candidate for the problem, given what you know about the data?

**Answer:

  • Gaussian Naive Bayes (GaussianNB)
  • An example of the use of Gaussian Naïve Bayes model in industry is using the model to identify stroke lesions from MRI imaging.1 Grffis et. Al. noted that by using this machine learning model there was less subjective bias in determining the extent of a lesion.
  • Naïve Bayes training and execution is fast with little resource utilization. With sufficient data, the Naïve Bayes algorithm is not sensitive to features that are irrelevant. This model performs the best with sufficient data that does not require features to be linked.
  • As for weaknesses, Naïve Bayes assumes that all features are independent. While this is beneficial if all variables are in fact independent this is rarely the case. For example, with the data set that we are using for this project we have the feature ‘reason’ which denotes why the student is taking the course. Students that are retaking the course are going to be more likely to have difficulties that might require academic intervention. This one feature should be weighted higher than ‘age’ or ‘free time’ features. Naïve Bayes may not work as well in cases like this where there is lots of data that should be considered collaboratively.
  • As noted above, Naïve Bayes is fast and resource efficient with easy to interpret results. Our data has a fairly large number of features that the Bayes algorithm will assess independently and then sum those probabilities to categorize whether a student is likely to require intervention. In our case, the data is relatively straightforward (students with more absences, less free time etc. are inherently less likely to be successful) which should lead to accurate results from the algorithm.

-

  • Support Vector Machines (SVM)
  • One way that Support Vector Machines are being used is to identify patients who are diagnosed with high-grade serous epithelial ovarian carcinoma (HGSOC) that are likely to be drug resistant. Sun et. Al. utilized samples that had been tested with 37 different immunology markers. The chemoresponse from those markers were then fed into a SVM classifier which successfully classified each patient as being likely to be chemo-resistant and also put the patients into high or low risk groups for progression. The results from the SVM algorithm ended up “outperforming all the other individual prognostic factors..”2 with an overall accuracy of 83.9%. 2
  • One major benefit of SVM is its ability to use the kernel trick. The kernel trick is useful in cases where the dataset is not linearly separable and allows the dataset to be transformed into a higher dimensional set that can be linearly separated. Additionally, SVM guarantees that the global optimum is always found as the classifier is set such that it is as far away from opposing data points as possible.
  • A major drawback in the use of SVM is how computationally intensive it is when there is large dataset or if data points are overlapping. In cases where computational power or training time is limited, another model should be chosen.
  • SVM is a good candidate for our model as there are limited overlapping data points and in this case computational power is not limited. The only thing that is a drawback from using SVM as opposed to a model like Naïve Bayes is that probability scores are not calculated. For our set of data there are many features that theoretically should have different weighting of importance as to whether a student needs academic support or not. As I understand it, SVM does not calculate probabilities on a per feature basis.

-

  • Logistic Regression
  • Another medicine based example for machine learning model use is the use of logistic regression to determine whether a patient is likely to be readmitted to a hospital following medical care. K. Zhu et al. used a California statewide de-identified medical database as a source for their data.3 They ran this data through a series of algorithms including logistic regression, SVM, random forest etc. They found that a conditional logistic regression model worked better than the other options by approximately 10%. The conditional logistic regression model also worked 20% than the standard logistic regression.3 This paper’s use case seems to be closely related to our use as it is a binary classification based on various data points that were collected for each patient.
  • Logistic regression models perform best when the data is simple and linear such as in the hospital readmission example above. Additionally, it works best when the output is to be binary i.e. a patient is going to be readmitted or not.
  • When a dataset does not have a single decision boundary logistic regression models are not very effectual. Logistic regression models also suffer from high bias which can lead to underfitting.
  • The logistic regression model works well for our dataset as our output is binary. Additionally, this model like Naïve Bayes, provides a probability for each feature. It also looks at the strength of each feature and weights it appropriately on its predictive usefulness.

-

  1. Griffis JC, Allendorfer JB, Szaflarski JP. Voxel-based Gaussian naive Bayes classification of ischemic stroke lesions in individual T1-weighted MRI scans. J Neurosci Methods 2016;257:97-108.
  2. Sun CY, Su TF, Li N, et al. A chemotherapy response classifier based on support vector machines for high-grade serous ovarian carcinoma. Oncotarget 2016;7:3245-54.
  3. Zhu K, Lou Z, Zhou J, Ballester N, Kong N, Parikh P. Predicting 30-day Hospital Readmission with Publicly Available Administrative Database. A Conditional Logistic Regression Modeling Approach. Methods Inf Med 2015;54:560-7. **

Setup

Run the code cell below to initialize three helper functions which you can use for training and testing the three supervised learning models you've chosen above. The functions are as follows:

  • train_classifier - takes as input a classifier and training data and fits the classifier to the data.
  • predict_labels - takes as input a fit classifier, features, and a target labeling and makes predictions using the F1 score.
  • train_predict - takes as input a classifier, and the training and testing data, and performs train_clasifier and predict_labels.
    • This function will report the F1 score for both the training and testing data separately.

In [10]:
def train_classifier(clf, X_train, y_train):
    ''' Fits a classifier to the training data. '''
    
    # Start the clock, train the classifier, then stop the clock
    start = time()
    clf.fit(X_train, y_train)
    end = time()
    
    # Print the results
    print "Trained model in {:.4f} seconds".format(end - start)

    
def predict_labels(clf, features, target):
    ''' Makes predictions using a fit classifier based on F1 score. '''
    
    # Start the clock, make predictions, then stop the clock
    start = time()
    y_pred = clf.predict(features)
    end = time()
    
    # Print and return results
    print "Made predictions in {:.4f} seconds.".format(end - start)
    return f1_score(target.values, y_pred, pos_label='yes')


def train_predict(clf, X_train, y_train, X_test, y_test):
    ''' Train and predict using a classifer based on F1 score. '''
    
    # Indicate the classifier and the training set size
    print "Training a {} using a training set size of {}. . .".format(clf.__class__.__name__, len(X_train))
    
    # Train the classifier
    train_classifier(clf, X_train, y_train)
    
    # Print the results of prediction for both training and testing
    print "F1 score for training set: {:.4f}.".format(predict_labels(clf, X_train, y_train))
    print "F1 score for test set: {:.4f}.".format(predict_labels(clf, X_test, y_test))

Implementation: Model Performance Metrics

With the predefined functions above, you will now import the three supervised learning models of your choice and run the train_predict function for each one. Remember that you will need to train and predict on each classifier for three different training set sizes: 100, 200, and 300. Hence, you should expect to have 9 different outputs below — 3 for each model using the varying training set sizes. In the following code cell, you will need to implement the following:

  • Import the three supervised learning models you've discussed in the previous section.
  • Initialize the three models and store them in clf_A, clf_B, and clf_C.
    • Use a random_state for each model you use, if provided.
    • Note: Use the default settings for each model — you will tune one specific model in a later section.
  • Create the different training set sizes to be used to train each model.
    • Do not reshuffle and resplit the data! The new training points should be drawn from X_train and y_train.
  • Fit each model with each training set size and make predictions on the test set (9 in total).
    Note: Three tables are provided after the following code cell which can be used to store your results.

In [13]:
# TODO: Import the three supervised learning models from sklearn
# from sklearn import model_A
from sklearn.naive_bayes import GaussianNB

# from sklearn import model_B
from sklearn import svm

# from sklearn import model_C
from sklearn.linear_model import LogisticRegression

# TODO: Initialize the three models
clf_A = GaussianNB()
clf_B = svm.SVC()
clf_C = LogisticRegression()

# TODO: Set up the training set sizes
X_train_100 = X_train[:100]
y_train_100 = y_train[:100]

X_train_200 = X_train[:200]
y_train_200 = y_train[:200]

X_train_300 = X_train[:300]
y_train_300 = y_train[:300]

# TODO: Execute the 'train_predict' function for each classifier and each training set size
# train_predict(clf, X_train, y_train, X_test, y_test)

# Gaussian Naive Bayes
train_predict(clf_A, X_train_100, y_train_100, X_test, y_test)
train_predict(clf_A, X_train_200, y_train_200, X_test, y_test)
train_predict(clf_A, X_train_300, y_train_300, X_test, y_test)

# Support Vector Machines
train_predict(clf_B, X_train_100, y_train_100, X_test, y_test)
train_predict(clf_B, X_train_200, y_train_200, X_test, y_test)
train_predict(clf_B, X_train_300, y_train_300, X_test, y_test)

# Logistic Regression
train_predict(clf_C, X_train_100, y_train_100, X_test, y_test)
train_predict(clf_C, X_train_200, y_train_200, X_test, y_test)
train_predict(clf_C, X_train_300, y_train_300, X_test, y_test)


Training a GaussianNB using a training set size of 100. . .
Trained model in 0.0010 seconds
Made predictions in 0.0006 seconds.
F1 score for training set: 0.8088.
Made predictions in 0.0006 seconds.
F1 score for test set: 0.6508.
Training a GaussianNB using a training set size of 200. . .
Trained model in 0.0015 seconds
Made predictions in 0.0007 seconds.
F1 score for training set: 0.7229.
Made predictions in 0.0005 seconds.
F1 score for test set: 0.6457.
Training a GaussianNB using a training set size of 300. . .
Trained model in 0.0010 seconds
Made predictions in 0.0014 seconds.
F1 score for training set: 0.6719.
Made predictions in 0.0003 seconds.
F1 score for test set: 0.6767.
Training a SVC using a training set size of 100. . .
Trained model in 0.2055 seconds
Made predictions in 0.0046 seconds.
F1 score for training set: 0.8372.
Made predictions in 0.0019 seconds.
F1 score for test set: 0.8485.
Training a SVC using a training set size of 200. . .
Trained model in 0.0058 seconds
Made predictions in 0.0067 seconds.
F1 score for training set: 0.8050.
Made predictions in 0.0016 seconds.
F1 score for test set: 0.8415.
Training a SVC using a training set size of 300. . .
Trained model in 0.0199 seconds
Made predictions in 0.0104 seconds.
F1 score for training set: 0.7879.
Made predictions in 0.0063 seconds.
F1 score for test set: 0.8485.
Training a LogisticRegression using a training set size of 100. . .
Trained model in 0.3132 seconds
Made predictions in 0.6265 seconds.
F1 score for training set: 0.8784.
Made predictions in 0.0005 seconds.
F1 score for test set: 0.7571.
Training a LogisticRegression using a training set size of 200. . .
Trained model in 0.0040 seconds
Made predictions in 0.1667 seconds.
F1 score for training set: 0.8084.
Made predictions in 0.0003 seconds.
F1 score for test set: 0.7391.
Training a LogisticRegression using a training set size of 300. . .
Trained model in 0.0052 seconds
Made predictions in 0.0132 seconds.
F1 score for training set: 0.7941.
Made predictions in 0.0005 seconds.
F1 score for test set: 0.8077.

Tabular Results

Edit the cell below to see how a table can be designed in Markdown. You can record your results from above in the tables provided.

Classifer 1 - Gaussian Naive Bayes

Training Set Size Training Time Prediction Time (test) F1 Score (train) F1 Score (test)
100 .0010s .0006s .8088 .6508
200 .0015s .0007s .7229 .6457
300 .0010s .0014s .6719 .6767

Classifer 2 - Support Vector Machines

Training Set Size Training Time Prediction Time (test) F1 Score (train) F1 Score (test)
100 .
200 EXAMPLE
300 EXAMPLE

Classifer 3 - ?

Training Set Size Training Time Prediction Time (test) F1 Score (train) F1 Score (test)
100
200
300

Choosing the Best Model

In this final section, you will choose from the three supervised learning models the best model to use on the student data. You will then perform a grid search optimization for the model over the entire training set (X_train and y_train) by tuning at least one parameter to improve upon the untuned model's F1 score.

Question 3 - Choosing the Best Model

Based on the experiments you performed earlier, in one to two paragraphs, explain to the board of supervisors what single model you chose as the best model. Which model is generally the most appropriate based on the available data, limited resources, cost, and performance?

Answer:

Question 4 - Model in Layman's Terms

In one to two paragraphs, explain to the board of directors in layman's terms how the final model chosen is supposed to work. Be sure that you are describing the major qualities of the model, such as how the model is trained and how the model makes a prediction. Avoid using advanced mathematical or technical jargon, such as describing equations or discussing the algorithm implementation.

Answer:

Implementation: Model Tuning

Fine tune the chosen model. Use grid search (GridSearchCV) with at least one important parameter tuned with at least 3 different values. You will need to use the entire training set for this. In the code cell below, you will need to implement the following:

  • Import sklearn.grid_search.GridSearchCV and sklearn.metrics.make_scorer.
  • Create a dictionary of parameters you wish to tune for the chosen model.
    • Example: parameters = {'parameter' : [list of values]}.
  • Initialize the classifier you've chosen and store it in clf.
  • Create the F1 scoring function using make_scorer and store it in f1_scorer.
    • Set the pos_label parameter to the correct value!
  • Perform grid search on the classifier clf using f1_scorer as the scoring method, and store it in grid_obj.
  • Fit the grid search object to the training data (X_train, y_train), and store it in grid_obj.

In [ ]:
# TODO: Import 'GridSearchCV' and 'make_scorer'

# TODO: Create the parameters list you wish to tune
parameters = None

# TODO: Initialize the classifier
clf = None

# TODO: Make an f1 scoring function using 'make_scorer' 
f1_scorer = None

# TODO: Perform grid search on the classifier using the f1_scorer as the scoring method
grid_obj = None

# TODO: Fit the grid search object to the training data and find the optimal parameters
grid_obj = None

# Get the estimator
clf = grid_obj.best_estimator_

# Report the final F1 score for training and testing after parameter tuning
print "Tuned model has a training F1 score of {:.4f}.".format(predict_labels(clf, X_train, y_train))
print "Tuned model has a testing F1 score of {:.4f}.".format(predict_labels(clf, X_test, y_test))

Question 5 - Final F1 Score

What is the final model's F1 score for training and testing? How does that score compare to the untuned model?

Answer:

Note: Once you have completed all of the code implementations and successfully answered each question above, you may finalize your work by exporting the iPython Notebook as an HTML document. You can do this by using the menu above and navigating to
File -> Download as -> HTML (.html). Include the finished document along with this notebook as your submission.