## Table of Contents

### Supervised Machine Learning: Regression Week 01 Quiz Answers

#### Quiz 01: Check for Understanding

Q1. Select the option that is the most INACCURATE regarding the definition of Machine Learning:

- Machine Learning allows computers to learn from data
- Machine Learning allows computers to infer predictions for new data
- Machine Learning is a subset of Artificial Intelligence
- Machine Learning is automated and requires no programming

Q2. This is the type of Machine Learning that uses both data with labeled outcomes and data without labeled outcomes:

- Supervised Machine Learning
- Unsupervised Machine Learning
- Mixed Machine Learning
- Semi-Supervised Machine Learning

Q3. Predicting total revenue, number of customers, and percentage of returning customers are examples of:

- classification
- regression

Q4. Predicting payment default, whether a transaction is fraudulent, and whether a customer will be part of the top 5% spenders on a given year, are examples of:

- classification
- regression

#### Quiz 02: Check for Understanding

Q1. Which statement about evaluating a Machine Learning model is the most accurate?

- Model selection involves choosing a model that minimizes the cost function.
- Model estimation involves choosing parameters that minimize the cost function.
- Model estimation involves choosing a cost function that can be compared across models.
- Model selection involves choosing modeling parameters that minimize in-sample validation error.

Q2. (True/False) The unadjusted value from estimating a linear regression model will almost always increase if more features are added.

- True
- False

Q3. (True/False) The Total Sum of Squares (TSS) can be used to select the best-fitting regression model.

- True
- False

Q4. (True/False) The Sum of Squared Errors (SSE) can be used to select the best-fitting regression model.

- True
- False

#### End of Module Quiz

Q1. You can use supervised machine learning for all of the following examples, EXCEPT:

- Segment customers by their demographics.
- Predict the number of customers that will visit a store on a given week.
- Predict the probability of a customer returning to a store.
- Interpret the main drivers that determine if a customer will return to a store.

Q2. The autocorrect on your phone is an example of:

- Unsupervised learning
- Supervised learning
- Semi-supervised learning
- Reinforcement learning

Q3. This is the type of Machine Learning that uses both data with labeled outcomes and data without labeled outcomes:

- Supervised Machine Learning
- Unsupervised Machine Learning
- Mixed Machine Learning
- Semi-Supervised Machine Learning

Q4. This option describes a way of turning a regression problem into a classification problem:

- Create a new variable that flags 1 for above a certain value and 0 otherwise
- Use outlier treatment
- Use missing value handling
- Create a new variable that uses autoencoding to transform a continuous outcome into categorical

Q5. This is the syntax you need to predict new data after you have trained a linear regression called *LR*:

- LR=predict(X_test)
- LR.predict(X_test)
- LR.predict(LR, X_test)
- predict(LR, X_test)

Q6. All of these options are useful error measures to compare regressions:

- SSE
- R squared
- TSS
- ROC index

Q7. (True/False) It is less concerning to treat a Machine Learning model as a black box for prediction purposes, compared to interpretation purposes:

- True
- False

### Supervised Machine Learning: Regression Week 02 Quiz Answers

#### Quiz 01: Check for Understanding

Q1. Another common term for the testing split is:

- Training split
- Validation split
- Corroboration split
- Cross validation split

Q2. Complete the following sentence: The training data is used to fit the model, while the test data is used to:

- measure the parameters and hyperparameters of the model
- tweak the model hyperparameters
- tweak the model parameters
- measure error and performance of the model

Q3. Select the option that has the syntax to obtain the data splits you will need to train a model having a test split that is a third the size of your available data.

- X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.5)
- X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.33)
- X_train, y_test = train_test_split(X, y, test_size=0.33)
- X_train, y_test = train_test_split(X, y, test_size=0.5)

#### Quiz 02: Check for Understanding

Q1. Which statement about K-fold cross-validation below is TRUE?

- Each subsample in K-fold cross-validation has at least
*k*observations. - Each subsample in K-fold cross-validation has at least
*k-1*observations. - Each of the
*k*subsamples in K-fold cross-validation is used as a training set. - Each of the
*k*subsamples in K-fold cross-validation is used as a test set.

Q2. (True/False) For a dataset with *M* observations and *N* features, Leave-one-out cross-validation is equivalent to k-fold cross-validation with *k* =*M-1* .

- True
- False

Q3. If a low-complexity model is underfitting during estimation, which of the following is MOST LIKELY true (holding the model constant)?

- K-fold cross-validation will still lead to underfitting, for any
*k*. - K-cross-validation with a small
*k*will reduce or eliminate underfitting. - K-fold cross-validation with a large
*k*will reduce or eliminate underfitting. - None of the above.

Q4. Which of the following statements about a high-complexity model in a linear regression setting is TRUE?

- Cross-validation with a small
*k*will reduce or eliminate overfitting. - A high variance of parameter estimates across cross-validation subsamples indicates likely overfitting.
- A low variance of parameter estimates across cross-validation subsamples indicates likely overfitting.
- Cross-validation with a large
*k*will reduce or eliminate overfitting.

#### Quiz 03: Check for Understanding

Q1. What is the main goal of adding polynomial features to a linear regression?

- Remove the linearity of the regression and turn it into a polynomial model.
- Capture the relation of the outcome with features of higher order.
- Increase the interpretability of a black box model.
- Ensure similar results across all folds when using K-fold cross validation.

Q2. What is the most common sklearn methods to add polynomial features to your data?

- polyFeat.add and polyFeat.transform
- polyFeat.add and polyFeat.fit
- polyFeat.fit and polyFeat.transform
- polyFeat.transform

#### End of Module Quiz

Q1. The main purpose of splitting your data into a training and test sets is:

- To improve accuracy
- To avoid overfitting
- To improve regularization
- To improve crossvalidation and overfitting

Q2. (True/False) For a dataset with *M* observations and *N* features, Stratified cross-validation is equivalent to k-fold cross-validation, where *k* =*N-1* .

- True
- False

Q3. (True/False) A linear regression model is being tested by cross-validation. Relative to K-fold cross-validation, stratified cross-validation (with the same *k* ) will likely increase the variance of estimated parameters.

- True
- False

Q4. In K-fold cross-validation, how will increasing *k* affect the variance (across subsamples) of estimated model parameters?

- Increasing
*k*will not affect the variance of estimated parameters. - Increasing
*k*will usually reduce the variance of estimated parameters. - Increasing
*k*will usually increase the variance of estimated parameters. - Increasing
*k*will increase the variance of estimated parameters if models are underfit, but reduce it if models are overfit.

### Supervised Machine Learning: Regression Week 03 Quiz Answers

#### Quiz 01: Check for Understanding

Q1. Which of the following statements about model complexity is TRUE?

- Higher model complexity leads to a lower chance of overfitting.
- Higher model complexity leads to a higher chance of overfitting.
- Reducing the number of features while adding feature interactions leads to a lower chance of overfitting.
- Reducing the number of features while adding feature interactions leads to a higher chance of overfitting.

Q2. Which of the following statements about model errors is TRUE?

- Underfitting is characterized by lower errors in both training and test samples.
- Underfitting is characterized by higher errors in both training and test samples.
- Underfitting is characterized by higher errors in training samples and lower errors in test samples.
- Underfitting is characterized by lower errors in training samples and higher errors in test samples.

Q3. Which of the following statements about regularization is TRUE?

- Regularization always reduces the number of selected features.
- Regularization increases the likelihood of overfitting relative to training data.
- Regularization decreases the likelihood of overfitting relative to training data.
- Regularization performs feature selection without a negative impact in the likelihood of overfitting relative to the training data.

Q4. BOTH Ridge regression and Lasso regression

- do not adjust the cost function used to estimate a model.
- add a term to the loss function proportional to a regularization parameter.
- add a term to the loss function proportional to the square of parameter coefficients.
- add a term to the loss function proportional to the absolute value of parameter coefficients.

Q5. Compared with Lasso regression (assuming similar implementation), Ridge regression is:

- less likely to overfit to training data.
- more likely to overfit to training data.
- less likely to set feature coefficients to zero.
- more likely to set feature coefficients to zero.

#### End of Module Quiz

Q1. (True/False) The variance of a model is determined by the degree of irreducible error.

- True
- False

Q2. (True/False) As more variables are added to a model, both its complexity and its variance generally increase.

- True
- False

Q3. (True/False) Model adjustments that decrease bias also decrease variance, leading to a bias-variance tradeoff.

- True
- False

Q4. Which of the following statements about scaling features prior to regularization is TRUE?

- The scale or features must be the same to implement L1 or L2 regularization.
- Features should rarely or never be scaled prior to implementing regularization.
- The larger a feature’s scale, the more likely its estimated impact will be influenced by regularization.
- The smaller a feature’s scale, the more likely its estimated impact will be influenced by regularization.

Q5. Which of the following statements about model complexity is TRUE?

- Higher model complexity leads to a lower chance of overfitting.
- Higher model complexity leads to a higher chance of overfitting.
- Reducing the number of features while adding feature interactions leads to a lower chance of overfitting.
- Reducing the number of features while adding feature interactions leads to a higher chance of overfitting.

Q6. (True/False) A model with high variance is characterized by sensitivity to small changes in input data.

- True
- False

Q7. Which of the following statements about Elastic Net regression is TRUE?

- Elastic Net combines L1 and L2 regularization.
- Elastic Net does not use L1 or L2 regularization.
- Elastic Net uses L2 regularization, as with Ridge regression.
- Elastic Net uses L1 regularization, as with Ridge regression.

##### Get All Course Quiz Answers of IBM Machine Learning Professional Certificate

Exploratory Data Analysis for Machine Learning Quiz Answers

Supervised Machine Learning: Regression Quiz Answers

Supervised Machine Learning: Classification Coursera Quiz Answers

Unsupervised Machine Learning Coursera Quiz Answers

Deep Learning and Reinforcement Learning Quiz Answers

Specialized Models: Time Series and Survival Analysis Quiz Answers