1. when might overfitting occur ANS >>> when the # of factors is close to or larger
than the # of data points causing the model to potentially fit too closely to random
effects
2. Why are simple models better than complex ones ANS >>> less data is
required; less chance of insignificant factors and easier to interpret
if it's good
3. what is forward selection ANS >>> we select the best new factor ith the current
and see enough (R^2, AIC, or p-value) add it to our model and fit the a certain
model w set of factors. Then at the end we remove factors that are
lower than threshold
he worst on a
4. what is backward elimination ANS >>> we start with all factors process over.
and find t supplied threshold (p = 0.15). If it is worse we remove it we move the
and start the We do that until we have the number of factors that we
want and then
factors lower than a second threshold (p = .05) and fit the model with all set of factors
5. what is stepwise regression ANS >>> it is a combination of
forward selec ward elimination. We can either start with all factors or
no factors and we remove or add a factor. As we go through the
procedure after ad factor and at the end we eliminate right away
factors that no longer a
https://www.stuvia.com/user/CULTIVATED
, ding each new
ppear.
6. what type of algorithms are stepwise selection? ANS >>> Greedy algorithms -
at each step they take one thing that looks best
7. what is LASSO ANS >>> a variable selection method where the coefficients are
olute
deter- mined by both minimizing the squared error and the sum of their absvalue not
being over a certain threshold t
8. How do you choose t in LASSO ANS >>> use the lasso approach with different
values of t and see which gives the best trade off
9. why do we have to scale the data for LASSO ANS >>> if we don't the
measure of the data will artificially affect how big the coefficients need to be
10. What is elastic net? ANS >>> A variable selection method that works by
coefficients
minimizing the squared error and constraining the combination of absolute values
of and their squares
11. what is a key difference between stepwise regresson and lasso regres-
sion ANS >>> If the data is not scaled, the coefficients can have artificially different
orders of magnitude, which means they'll have unbalanced effects on the lasso
constraint.
12. Why doesn't Ridge Regression perform variable selection? ANS >>>
The coeffi- cients values are squared so they go closer to zero or regularizes
them
13. What are the pros and cons of Greedy Algorithms (Forward selection,
stepwise elimination, stepwise regression) ANS >>> Good for initial analysis
but often don't perform as well on other data because they fit more to random
effects than you'd like and appear to have a better fit
14. What are the pros and cons of LASSO and elastic net ANS >>> They are
slower but help make models that make better predictions
https://www.stuvia.com/user/CULTIVATED
, 15. Which two methods does elastic net look like it combines and what are the
downsides from it? ANS >>> Ridge Regression and LASSO.
Advantages ANS >>> variable selection from LASSO and Predictive benefits of
LASSO.
Disadvantages ANS >>> Arbitrarily rules out some correlated variables like LASSO
(don't know which one that is left out should be); Underestimates coefficients of ve
predictive variables like Ridge Regresison
16. What are some downsides of surveys? ANS >>> Even if you what appears
to be a representative sample in simple ways, maybe it isn't in more complex
ways.
17. If we're testing to see whether red cars sell for higher prices than blue cars,
we need to account for the type and age of the cars in our data set. This is
called ANS >>> ANS >>> Controlling
18. what is a blocking factor ANS >>> a source of variability that is not of primary
interest to the experimenter
19. what is an example of a blocking factor ANS >>> The type of car, sports car
or family car, is a blocking factor that it could account for some of the difference
between red cars and blue cars. Because sports cars are more likely to be red; if
we account for the difference, we can reduce the variability in our estimates
20. Under what conditions should you run A/B tests ANS >>> When you can
collect data quickly. When the data is representative and the amount of data is
small compared to the whole population
21. Do you have to decide the sample size ahead of time for A/B tests ANS >>>
no, and we can run the hypothesis test anytime we want
22. What is full factorial design ANS >>> you test every combination and then use
ANOVA to determine importance of each factor
23. What is fractional factorial design ANS >>> when you test a subset of the
entire set of combinations
24. What is a balanced design? ANS >>> You test each choice the f times and
same # o each pair of choices the same # of times
25. When is regression effective work well to determine important factors?
ANS >>> If there aren't significant interactions between the factors.
26. what is exploration? ANS >>> focusing on getting more information; in this
case, to determine with more certainty which ad is really the best
27. what is exploitation ANS >>> we're focused on getting immediate value; in this
example, to show the add that seems to be doing best so far, because it seems to
be most likely to be clicked.
https://www.stuvia.com/user/CULTIVATED
than the # of data points causing the model to potentially fit too closely to random
effects
2. Why are simple models better than complex ones ANS >>> less data is
required; less chance of insignificant factors and easier to interpret
if it's good
3. what is forward selection ANS >>> we select the best new factor ith the current
and see enough (R^2, AIC, or p-value) add it to our model and fit the a certain
model w set of factors. Then at the end we remove factors that are
lower than threshold
he worst on a
4. what is backward elimination ANS >>> we start with all factors process over.
and find t supplied threshold (p = 0.15). If it is worse we remove it we move the
and start the We do that until we have the number of factors that we
want and then
factors lower than a second threshold (p = .05) and fit the model with all set of factors
5. what is stepwise regression ANS >>> it is a combination of
forward selec ward elimination. We can either start with all factors or
no factors and we remove or add a factor. As we go through the
procedure after ad factor and at the end we eliminate right away
factors that no longer a
https://www.stuvia.com/user/CULTIVATED
, ding each new
ppear.
6. what type of algorithms are stepwise selection? ANS >>> Greedy algorithms -
at each step they take one thing that looks best
7. what is LASSO ANS >>> a variable selection method where the coefficients are
olute
deter- mined by both minimizing the squared error and the sum of their absvalue not
being over a certain threshold t
8. How do you choose t in LASSO ANS >>> use the lasso approach with different
values of t and see which gives the best trade off
9. why do we have to scale the data for LASSO ANS >>> if we don't the
measure of the data will artificially affect how big the coefficients need to be
10. What is elastic net? ANS >>> A variable selection method that works by
coefficients
minimizing the squared error and constraining the combination of absolute values
of and their squares
11. what is a key difference between stepwise regresson and lasso regres-
sion ANS >>> If the data is not scaled, the coefficients can have artificially different
orders of magnitude, which means they'll have unbalanced effects on the lasso
constraint.
12. Why doesn't Ridge Regression perform variable selection? ANS >>>
The coeffi- cients values are squared so they go closer to zero or regularizes
them
13. What are the pros and cons of Greedy Algorithms (Forward selection,
stepwise elimination, stepwise regression) ANS >>> Good for initial analysis
but often don't perform as well on other data because they fit more to random
effects than you'd like and appear to have a better fit
14. What are the pros and cons of LASSO and elastic net ANS >>> They are
slower but help make models that make better predictions
https://www.stuvia.com/user/CULTIVATED
, 15. Which two methods does elastic net look like it combines and what are the
downsides from it? ANS >>> Ridge Regression and LASSO.
Advantages ANS >>> variable selection from LASSO and Predictive benefits of
LASSO.
Disadvantages ANS >>> Arbitrarily rules out some correlated variables like LASSO
(don't know which one that is left out should be); Underestimates coefficients of ve
predictive variables like Ridge Regresison
16. What are some downsides of surveys? ANS >>> Even if you what appears
to be a representative sample in simple ways, maybe it isn't in more complex
ways.
17. If we're testing to see whether red cars sell for higher prices than blue cars,
we need to account for the type and age of the cars in our data set. This is
called ANS >>> ANS >>> Controlling
18. what is a blocking factor ANS >>> a source of variability that is not of primary
interest to the experimenter
19. what is an example of a blocking factor ANS >>> The type of car, sports car
or family car, is a blocking factor that it could account for some of the difference
between red cars and blue cars. Because sports cars are more likely to be red; if
we account for the difference, we can reduce the variability in our estimates
20. Under what conditions should you run A/B tests ANS >>> When you can
collect data quickly. When the data is representative and the amount of data is
small compared to the whole population
21. Do you have to decide the sample size ahead of time for A/B tests ANS >>>
no, and we can run the hypothesis test anytime we want
22. What is full factorial design ANS >>> you test every combination and then use
ANOVA to determine importance of each factor
23. What is fractional factorial design ANS >>> when you test a subset of the
entire set of combinations
24. What is a balanced design? ANS >>> You test each choice the f times and
same # o each pair of choices the same # of times
25. When is regression effective work well to determine important factors?
ANS >>> If there aren't significant interactions between the factors.
26. what is exploration? ANS >>> focusing on getting more information; in this
case, to determine with more certainty which ad is really the best
27. what is exploitation ANS >>> we're focused on getting immediate value; in this
example, to show the add that seems to be doing best so far, because it seems to
be most likely to be clicked.
https://www.stuvia.com/user/CULTIVATED
