MGSC Exam 3 questions with verified answers
>Baicc["SIDNEY_CROSBY"]SIDNEY_CROSBY
0.4131174 > exp(Baicc["SIDNEY_CROSBY"])SIDNEY_CROSBY
1.511523
Which of the following statements is correct? Ans✓✓✓ The odds of getting a
home goal is increased 50% when Crosby is on the ice
A 95% confidence interval implies Ans✓✓✓ the parameter has a 95% chance of
being within the range of values.
A high p-value indicates Ans✓✓✓ lack of statistical significance
common result if null is true.
A lasso path plot should be read Ans✓✓✓ from right to left
A low p-value indicates Ans✓✓✓ statistical significance
you observed a rare result if the null were true.
Adding so many regression predictors that you tune your model to the training
data and it doesn't work well on new data is known as Ans✓✓✓ overfit (high
variance)
, All that matters is Ans✓✓✓ out-of-sample deviance (R-squared), can be negative
As you move from the biggest to the smallest lambda in regularization, the model
has Ans✓✓✓ more predictors
Bootstrap sample Ans✓✓✓ Original ($t0), bias, standard error, and $t
Correct for bias Ans✓✓✓ use bootstrap to estimate the bias and adjust estimates
Dependencies in the data often lead to an Ans✓✓✓ underestimate of
uncertainty
Family= "binomial" in a code means its logistic or linear? Ans✓✓✓ Logistic lasso
Greedy search Ans✓✓✓ just looking at local area of where you are now.
How do you estimate the standard error of the sampling distribution using a
bootstrap? Ans✓✓✓ calculate the standard deviation of the bootstrap estimates
How many coefficients are in the model at the end of the lasso path? Ans✓✓✓
2189 left side
Read the plot from right to left. The lambda decreases which means the penalty is
the lowest at the left side of the plot and when the penalty is low, more
coefficients enter the model.
>Baicc["SIDNEY_CROSBY"]SIDNEY_CROSBY
0.4131174 > exp(Baicc["SIDNEY_CROSBY"])SIDNEY_CROSBY
1.511523
Which of the following statements is correct? Ans✓✓✓ The odds of getting a
home goal is increased 50% when Crosby is on the ice
A 95% confidence interval implies Ans✓✓✓ the parameter has a 95% chance of
being within the range of values.
A high p-value indicates Ans✓✓✓ lack of statistical significance
common result if null is true.
A lasso path plot should be read Ans✓✓✓ from right to left
A low p-value indicates Ans✓✓✓ statistical significance
you observed a rare result if the null were true.
Adding so many regression predictors that you tune your model to the training
data and it doesn't work well on new data is known as Ans✓✓✓ overfit (high
variance)
, All that matters is Ans✓✓✓ out-of-sample deviance (R-squared), can be negative
As you move from the biggest to the smallest lambda in regularization, the model
has Ans✓✓✓ more predictors
Bootstrap sample Ans✓✓✓ Original ($t0), bias, standard error, and $t
Correct for bias Ans✓✓✓ use bootstrap to estimate the bias and adjust estimates
Dependencies in the data often lead to an Ans✓✓✓ underestimate of
uncertainty
Family= "binomial" in a code means its logistic or linear? Ans✓✓✓ Logistic lasso
Greedy search Ans✓✓✓ just looking at local area of where you are now.
How do you estimate the standard error of the sampling distribution using a
bootstrap? Ans✓✓✓ calculate the standard deviation of the bootstrap estimates
How many coefficients are in the model at the end of the lasso path? Ans✓✓✓
2189 left side
Read the plot from right to left. The lambda decreases which means the penalty is
the lowest at the left side of the plot and when the penalty is low, more
coefficients enter the model.
