Reading Quiz #5 (2022W2)

Started: Feb 13 at 8p.m.

Quiz Instructions
To prepare for this quiz, please read sections 6.1-6.2 (inclusive) in the Kleinberg + Tardos textbook.

The goal of this quiz is to lightly assess a first quick reading of these resources to prepare for class. You
should definitely return to this material for a more thorough read to solidify your learning and prepare for
assignments and exams.

Note that you are limited to 3 attempts for this quiz.

Best of luck! :-)

Question 1 1 pts

Consider the Weighted Interval Scheduling Problem (WISP) as defined in section 6.1 of
the textbook. Consider two potential greedy algorithms that a CPSC 320 student might
be tempted to apply to solve it:

GreedyEarliestFinish is our name for exactly the algorithm you learned about in
section 4.1 of the textbook; it chooses first the interval with the earliest finish time,
discards all intervals that conflict with it, and then repeats for the interval with the next
earliest finish time, etc.
GreedyHighestValue is one of the natural greedy algorithms you might expect for
the Weighted Interval Scheduling problem: first choose the interval with the greatest
weight/value, discard all intervals conflicting with it, and then iterate on the remaining
intervals.

Now, consider the following non-trivial, but still manageable, instance of WISP, taken
directly from the same textbook section (with a single minor change, to make it a little
more interesting):
For this question, please match each algorithm on the left to the value of the solution it
would find for this instance. (For the "Optimal" choice, it doesn't matter which algorithm
generated the solution; we're only interested in the greatest value possibly achievable on
this particular instance.)

GreedyEarliestFinish [ Choose ]

GreedyHighestValue [ Choose ]

Optimal [ Choose ]

Question 2 1 pts

Which of the following insights are used in determining the recurrence relation for the
Weighted Interval Scheduling Problem (WISP)? Choose all that apply.

In the choices below, assume that is the total number of intervals in an instance of
WISP, and that the intervals have already been sorted in order of non-decreasing finish
time, and is the index of some arbitrary interval, with .
p(j) is defined as in Section 6.1 of the textbook with a typo correction to the book's
English description: it gives the index of the rightmost interval that ends before j begins.

All intervals in the optimal solution for the first j intervals must be included in the optimal
solution for the first j+1 intervals

The optimal schedule that includes the nth interval adds its value to the optimal solution for
the first p(n) intervals.

If there are exactly two choices for whether the jth interval is included in the optimal
solution, the optimal solution must include the lowest in value of these choices

For any j between 1 and n, if we know the optimal scheduling for the first j-1 intervals,
finding the optimal solution for the first j intervals amounts to making a binary choice.

The original problem instance can be expressed as a subproblem of itself

Question 3 1 pts

Below is a table containing all pertinent information from the execution of the Dynamic
Programming (DP) algorithm for WISP on a particular instance containing 6 intervals.
The columns of the table are labelled according to the following legend:

The j column lists the indices of each interval. The intervals have been numbered in
order of non-decreasing finish time. (The zeroth index is a placeholder to make the
recurrence relation nicer.)
v_j gives the weight/value of the jth interval.
p(j) is defined as in Section 6.1 of the textbook with a typo correction to the
book's English description: it gives the index of the rightmost interval that ends
before j begins.
M[j] is the memo table populated by the DP algorithm; it contains the value of the
optimal solution considering only the first j intervals.
Using this table, along with the recurrence relation for WISP given in Section 6.1 of the
textbook (Equation 6.1), reconstruct the solution found by the DP algorithm, by
selecting the index ("j") of EACH interval that gets scheduled in the optimal (highest-
weight) scheduling.

Question 4 1 pts

Which of these differs between a recursive, memoized implementation of DP, vs a

(Elmer Fudd is not the right answer, but you can Google Seawch
(https://www.google.ca/search?hl=xx-elmer) for him.)

Whether we ask for the solution to a particular subproblem more than once.

Whethuh Elmuh Fudd was taught it in grade school.

By the time we consider a particular problem, whether the solutions to its subproblems have
already been computed.

Whether we compute the solution to a particular subproblem more than once.

Not saved Submit Quiz