is constructed by the individual optimal objectives value

Metaheuristics need to balance two important objectives: intensification and diversification:

Diversification explores the search space

Pareto front includes all efficient solutions. These are solutions which:

cant improve in any obj without degrading in at least 1other

A local search heuristic is based on successive moves over a neighborhood structure:

Is highly dependent on the starting solution.

A Memetic Algorithm is a hybridization of GA with

local search after crossover and mutation

A fundamental difference between multiple (MO) and single (SO) objective optimization is

Confliting MO no optimal solut while SO might have 1 or more

Should a metaheuristic be used to a specific problem?

Depends on the complexity of the problem

The strategy to create the new generation

The nadir objective vector represents

Upper bounds of each min objective in the pareto-optimal set

a point-centered iterative search

Tabu search incorporated intelligence:

Tabu-lists are introduced to prevent cycle back

One can say that an heuristic is an algorithm that

Builds and/or improve a solution

Has more variability than others industries

The basics concepts to problem solving are

representation, objective and evaluation

a limited quantity of inf related to the last movements

In each iteration of a tabu search algorithm, the solution moves to the best

not forbidden solution in the neighborhood

The most critical step in the design of a tabu-search algorithm is:

Search space and neighborhood structure

About parameter calibration and computer experiences

large sets of instances should be used to ensure robustness

A greedy heuristics builds a solution in a series of steps

in every step chooses the best available decisions

An heuristic method is a/an

procedures to discover good but near-optimal solutions

The best improvement strategy for a local search procedure

Needs to explore all the neighborhood before moving

studbud

CIO2

30 termsStudBud Administration27 plays

"RumoAo20

Flashcards30 cardsFPractice quizPHost liveL

Questions 30

Answers

1
About evolution in genetic algorithms
there is no evidence of evolution through generation
is mostly obtained through mutation
survival of the best fitted
decreasing fitness through generations
2
The ideal vector:
is constructed by the individual optimal objectives value
provides lower bounds on each max obj to normalize obj value
corresponds to a feasible solution
includes components better than the individual obj values
3
Metaheuristics need to balance two important objectives: intensification and diversification:
Diversification explores the search space
Diversification is obtained through local search
Diversification explores the best solution found so far
Intensification is a global search
4
Pareto front includes all efficient solutions. These are solutions which:
are dominated by at least one non-dominated solution
are individual optimal solutions of each obj function
cant improve in any obj without degrading in at least 1other
dominate at least one other solution
5
Prof Derya Demirtas is originally from
Canada
The Netherlands
Turkey
Russia
6
A local search heuristic is based on successive moves over a neighborhood structure:
Applied to a population of solutions
Is a complex algorithm to design, implement and apply
Neighborhood is independent of the problem solving situation
Is highly dependent on the starting solution.
7
A Memetic Algorithm is a hybridization of GA with
tabu search after selection and replacement
local search after crossover and mutation
constructive search heuristics after mutations
greedy algorithm for the initial generation
8
A fundamental difference between multiple (MO) and single (SO) objective optimization is
None since a MO can be easily transformed into SO
Several optimal solutions for MO and only one for SO
Confliting MO no optimal solut while SO might have 1 or more
MO are more adjusted to real-world problems than SO
9
BLP formulation of conditional constraints: “if site 1 is open, then sites 2 and 3 must open”
y2+y3 ≤ y1
y2+y3 ≥ y1
y2 ≥ y1 and y3 ≥ y1
y2 ≤ y1 and y3 ≤ y1
10
A genetic algorithm is:
An iterative and single point process
Is based on artificial selection
was developed in the yearly wears of XX century
is bio-inspired
11
Should a metaheuristic be used to a specific problem?
Depends on the complexity of the problem
Depends on the available instances
A metaheuristic is always too complex to be implemented
A metaheuristic fits well many problems thus should be used
12
Replacement is:
The strategy to create the new generation
Replaces all current population by the new chromosomes
problem-specific
Eliminates duplicates to avoid algorithm convergence
13
The nadir objective vector represents
A feasible solution
The worst possible function values in the search space
Upper bounds of each min objective in the pareto-optimal set
an infeasible solution
14
A tabu search is
a metaheuristic population based
a chance-centered algorithm
a point-centered iterative search
based on several neighborhood structures
15
Tabu search incorporated intelligence:
Non-improvement moves are allowed to prevent cycle back
Tabu-lists are introduced to prevent cycle back
Tabu lists are introduced to allow diversification
Non-improvement moves are allowed to diversify the search
16
One can say that an heuristic is an algorithm that
Builds and/or improve a solution
Builds and improves a solution
Builds a solution following greedy rules
Gives an optimal solution
17
Healthcare is:
A system where activities are always preemptive
Has a standard product that applies to almost all customers
A business like all others
Has more variability than others industries
18
The basics concepts to problem solving are
size of the search space
programming, goal and computacional time
feasibility and optimality
representation, objective and evaluation
19
A tabu list includes
all the attractive components
all the already visited solutions
a limited quantity of inf related to the last movements
the elements that the local search might use
20
In each iteration of a tabu search algorithm, the solution moves to the best
improvement not forbidden solution in the neighborhood
not forbidden solution in the neighborhood
improvement solution in the neighborhood
neighbor solution
21
The most critical step in the design of a tabu-search algorithm is:
Search space and neighborhood structure
termination criteria
Tabu-moves representation
How long a movement or solution should remain tabu
22
Set-covering model is used on:
Portfolio optimization
Facility location
Distribution routes
Production scheduling
23
About parameter calibration and computer experiences
Only the real instances of the case study should be used
aim to find good parameters to the instances at hand
several equal complexity and size instances should be used
large sets of instances should be used to ensure robustness
24
A greedy heuristics builds a solution in a series of steps
is it an example of an improvement heuristic
in every step chooses the best available decisions
is it an example of a compound heuristic
returns the optimal solution overall
25
An heuristic method is a/an
procedures to discover good but near-optimal solutions
structured short-cut to discover an optimal solution
An exhaustive and complete search of the possible solutions
intelligent way to generate all feasible solutions
26
Alvin Roth worked on a modern healthcare operations management project:
Kidney exchange
Blood donation planning and inventory management
Defibrillators location
Epidemiology
27
In genetic algorithms, diversification is obtained mainly through
Selection
Evaluation
Representation
Mutation
28
Aspiration Criterion is
Frequency based memory
Memory with attractive componentes
Memory to ensure diversity in the search
Device to cancel tabus
29
The best improvement strategy for a local search procedure
Needs to explore all the neighborhood before moving
Is more robust because repeats for ≠ initial solutions
At each iteration is quicker than the first improvemen
Moves as soon as a better solution is obtained
30
Select a cut to the constraint: x1+ 2x2+ 4x3 + 9x4 + 10x5 + 20x6 ≤ 20
x5+ x6 ≤ 1
x1+ x2+ x3 + x4 + x5 + x6 ≤ 1
x1+ x2+ x3 + x4 ≤ 3
x1+ x2+ x3 + x4 + x5 + x6 ≤ 6

CIO2 Flashcards | StudBud