DESCRIPTION

Students:  Please note instructions in paper copy of quiz.  Save your
files often, make sure OMSI fills your entire screen at all times, etc.

A question may not fully fit into your OMSI question box, which is not
scrollable.  You can try adjusting the relative size of the question and
answer boxes, but remember that clicking CopyQtoA will copy the entire
question box to the answer box.

QUESTION

(20 pts) An alternative term for "embedding," common in economics and
the social sciences is _______________________.

QUESTION

(20 pts) Answer "smaller" or "larger":  In CART, the larger the number of nodes,
the ______________ the variance.  In k-NN, the larger the value of k,
the ___________ the variance.  In NMF, the larger the rank, the
___________ the variance.  

QUESTION

(20 pts) In the paper by Lee and Seung, what term do they use for the l2
norm?

QUESTION -ext .R -run 'Rscript omsi_answer4.R'

(20 pts) In the InstEval data, use CART to predict rating from all other
columns except the user and item IDs.  Your prediction will be for a
case in which the course is a service course in department 3, and
studage and lectage are 4 and 1, respectively.  Note:  Check first which
columns are R factors; these must be accessed via quotation marks, i.e.
as character strings.  Use the default values of the tuning parameters.
Make sure to use print() to output your answer!

QUESTION -ext .R -run 'Rscript omsi_answer5.R'

(20 pts) (This question is independent of Question 4.) Suppose that in
the InstEval data we will use an embedding of mean user rating in place
of the 's' column.  (We will not use the 'd' column at all.) Write code
that will generate a new data frame ie_emb that follows this plan.  Your
code must use formUserData(), and must end with

print(ie_emb[1,])