1|Page
Mid1: ISYE 6501 Midterm Quiz 1 2026-2027 BANK QUESTIONS
WITH DETAILED VERIFIED ANSWERS EXAM QUESTIONS
WILL COME FROM HERE (100% CORRECT ANSWERS A+
GRADED
1. In the context of statistical learning, what is the primary goal of
supervised learning?
A) To find hidden structures in unlabeled data
B) To maximize the reward in a dynamic environment
C) To model the relationship between a response variable and a set of
predictors
D) To reduce the dimensionality of the data while preserving variance
Answer: C
Explanation: Supervised learning involves building a model that relates
a response (dependent variable) to one or more predictors
(independent variables) using labeled training data.
2. Which of the following is a typical example of a classification
problem?
A) Predicting the price of a stock
B) Grouping customers into market segments
C) Diagnosing whether a patient has a disease
,2|Page
D) Estimating the number of units sold
Answer: C
Explanation: Classification problems involve predicting a categorical
outcome, such as the presence or absence of a disease.
3. The bias-variance trade-off refers to the balance between which two
sources of error?
A) Type I error and Type II error
B) Training error and test error
C) Error from simplifying assumptions and error from sensitivity to data
fluctuations
D) Systematic error and random error in instrumentation
Answer: C
Explanation: Bias refers to error from overly simplistic assumptions in
the learning algorithm, while variance refers to error from excessive
sensitivity to small fluctuations in the training set.
4. In a regression setting, what does the irreducible error represent?
A) The error that can be eliminated by adding more predictors
B) The error due to the model's inability to capture the true relationship
C) The variability in the response that cannot be explained by the
predictors
D) The error introduced by the optimization algorithm
Answer: C
,3|Page
Explanation: The irreducible error is the noise inherent in the system,
representing the lower bound on the expected prediction error that
cannot be eliminated by any model.
5. What does a training Mean Squared Error (MSE) that is much lower
than the test MSE typically indicate?
A) The model has high bias
B) The model is underfitting
C) The model is overfitting
D) The irreducible error is zero
Answer: C
Explanation: This situation indicates overfitting, where the model has
learned the random noise in the training data instead of just the
underlying signal, leading to poor generalization.
6. In a linear regression model Y = β0 + β1 X + ε, what does the error
term ε represent?
A) The predicted value of the response
B) The deviation of the predicted value from the mean
C) The random noise or deviation of the true response from the linear
relationship
D) The intercept of the regression line
Answer: C
, 4|Page
Explanation: The error term captures all factors that influence the
response Y other than the predictor X, including random noise and
unmodeled variables.
7. Which method is used to estimate the coefficients in a standard
linear regression model?
A) Maximum a posteriori estimation
B) Gradient boosting
C) Ordinary Least Squares
D) K-means clustering
Answer: C
Explanation: Ordinary Least Squares (OLS) is the standard method that
minimizes the sum of squared differences between the observed and
predicted values to estimate the coefficients.
8. How does a large value of the tuning parameter λ in ridge regression
affect model complexity compared to ordinary least squares?
A) It increases complexity by adding polynomial terms
B) It decreases complexity by shrinking coefficients toward zero
C) It has no effect on complexity but improves interpretability
D) It increases complexity by including more variables
Answer: B
Explanation: Ridge regression adds a penalty (λ times the sum of
squared coefficients) to the least squares objective. A larger λ shrinks
Mid1: ISYE 6501 Midterm Quiz 1 2026-2027 BANK QUESTIONS
WITH DETAILED VERIFIED ANSWERS EXAM QUESTIONS
WILL COME FROM HERE (100% CORRECT ANSWERS A+
GRADED
1. In the context of statistical learning, what is the primary goal of
supervised learning?
A) To find hidden structures in unlabeled data
B) To maximize the reward in a dynamic environment
C) To model the relationship between a response variable and a set of
predictors
D) To reduce the dimensionality of the data while preserving variance
Answer: C
Explanation: Supervised learning involves building a model that relates
a response (dependent variable) to one or more predictors
(independent variables) using labeled training data.
2. Which of the following is a typical example of a classification
problem?
A) Predicting the price of a stock
B) Grouping customers into market segments
C) Diagnosing whether a patient has a disease
,2|Page
D) Estimating the number of units sold
Answer: C
Explanation: Classification problems involve predicting a categorical
outcome, such as the presence or absence of a disease.
3. The bias-variance trade-off refers to the balance between which two
sources of error?
A) Type I error and Type II error
B) Training error and test error
C) Error from simplifying assumptions and error from sensitivity to data
fluctuations
D) Systematic error and random error in instrumentation
Answer: C
Explanation: Bias refers to error from overly simplistic assumptions in
the learning algorithm, while variance refers to error from excessive
sensitivity to small fluctuations in the training set.
4. In a regression setting, what does the irreducible error represent?
A) The error that can be eliminated by adding more predictors
B) The error due to the model's inability to capture the true relationship
C) The variability in the response that cannot be explained by the
predictors
D) The error introduced by the optimization algorithm
Answer: C
,3|Page
Explanation: The irreducible error is the noise inherent in the system,
representing the lower bound on the expected prediction error that
cannot be eliminated by any model.
5. What does a training Mean Squared Error (MSE) that is much lower
than the test MSE typically indicate?
A) The model has high bias
B) The model is underfitting
C) The model is overfitting
D) The irreducible error is zero
Answer: C
Explanation: This situation indicates overfitting, where the model has
learned the random noise in the training data instead of just the
underlying signal, leading to poor generalization.
6. In a linear regression model Y = β0 + β1 X + ε, what does the error
term ε represent?
A) The predicted value of the response
B) The deviation of the predicted value from the mean
C) The random noise or deviation of the true response from the linear
relationship
D) The intercept of the regression line
Answer: C
, 4|Page
Explanation: The error term captures all factors that influence the
response Y other than the predictor X, including random noise and
unmodeled variables.
7. Which method is used to estimate the coefficients in a standard
linear regression model?
A) Maximum a posteriori estimation
B) Gradient boosting
C) Ordinary Least Squares
D) K-means clustering
Answer: C
Explanation: Ordinary Least Squares (OLS) is the standard method that
minimizes the sum of squared differences between the observed and
predicted values to estimate the coefficients.
8. How does a large value of the tuning parameter λ in ridge regression
affect model complexity compared to ordinary least squares?
A) It increases complexity by adding polynomial terms
B) It decreases complexity by shrinking coefficients toward zero
C) It has no effect on complexity but improves interpretability
D) It increases complexity by including more variables
Answer: B
Explanation: Ridge regression adds a penalty (λ times the sum of
squared coefficients) to the least squares objective. A larger λ shrinks
