Brainscape's Knowledge GenomeTM

Browse over 1 million classes created by top students, professors, publishers, and experts.


See full index

Swarm Intelligence > reviewing randomly > Flashcards

reviewing randomly Flashcards

1
Q

What does it mean for a system to be self-protecting?

A

A system is self-protecting if it continuously maintains the safety property in the presence of external actions with “malicious intent”.
E.g., fail-safe in the presence of attacks on peer-to-peer networks, etc.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
2
Q

What does it mean for a system to be self-explanatory?

A

The system control explains why it behaved in an observed manner

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
3
Q

What does it mean for a system to be self-healing?

A

A system is self-healing with respect to a set C of external (bad) events if the occurrence of events from C violates the safety property of the system just for a short time.
E.g., removing nodes from peer-to-peer networks

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
4
Q

What does it mean for a system to be self-optimizing?

A

A system is self-optimizing if – starting from any initial configuration – it is able to optimize the value of a predetermined objective function over the global state.
E.g., minimization of energy consumption

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
5
Q

Name 3 reasons for the increasing complexity of technical systems.

A

Miniaturization
Interconnection
embedding

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
6
Q

What is the liveness-property?

A

“good things happen sooner or later

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
7
Q

What is the main idea behind systems with self-*-properties?

A

In order to enable the correctness of such a system despite increasing complexity, it becomes more and more important that some processes within the system run automatically, i.e., without human intervention.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
8
Q

What does it mean for a system to be self-configuring?

A

A system is self-configuring with respect to a set of actions if it is able to change its own configuration in order to restore or improve the safety property.
adaptation of
the hardware configuration in cloud computing depending on the load
e.g. adaptation of
the hardware configuration in cloud computing depending on the load

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
9
Q

How does the convergence behavior in Particle Swarm Optimization (PSO) vary with different actual eigenvalues of the dynamic matrix ‘A’ whose properties determine the time behavior of the particle?

A

Complex values –>harmonic oscillations
at least one of the eigenvalues, whether real or complex, has a negative real part –> zigzagging

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
10
Q

Name 5 self-*-properties

A

Self-configuring
self-optimizing
self-healing
self-explanatory
self-protecting

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
11
Q

what is the safety property?

A

bad things never happen

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
12
Q

What are the advantages of systems that have self-*-properties?

A

flexible
robust against failures
self-optimizing

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
13
Q

What are the disadvantages of systems that have self-*-properties?

A

can make mistakes
very long training times, very long times for reconfiguration
unauthorized interference possible

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
14
Q

What are the patterns of organization used in systems with self-*-properties?

A

emergence
autonomy (state of self-sufficiency, independence, self-government)
federation (cooperation of several (sub-)systems)
self-organization

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
15
Q

Define the term Emergence

A

Emergence is a phenomenon characterized by the interaction of many components and the absence of central control and explicitly predetermined patterns.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
16
Q

What are the typical phenomena for emergence? Give several examples of emergence behavior.

A

unpredictability
irreducibility
-oscillating circuit, especially the phenomenon of resonance
-stability of living systems against environmental influence
-performing arts, especially when abstract
web search (significance of a page is derived from link structure)

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
17
Q

Which research questions were discussed in the lecture in relation to the term Emergence and which algorithm was used?

A

(better and) quantitative understanding of natural phenomena
metrics for the assessment of self-organization and emergence phenomena
system architectures: Observer/Controller architectures
security: the self-evolution of the OC system must prevent misbehavior and misdevelopment
inclusion of a-priori knowledge
cognitive ability (perceptiveness) autonomy and user interaction
self-explanation

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
18
Q

What is PSO algorithm? What is it for?

A

a method for finding the minimum of a continuous function.
works in an evolutionary manner,
i.e., already obtained solution candidates are improved (step by step).
The progress takes place in generations (iterations), where the improvement happens by imitating and learning from other individuals in the population.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
19
Q

How is the continuous function given in the PSO?

A
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
20
Q

Which 2 different heuristics are needed for good results at PSO?

A

Exploration: the ability of the swarm to search the search space “completely.” Important especially at the starting phase of the algorithm.
Exploitation: Ability of the swarm to “find near good solutions even better solutions”. This is important, especially towards the end phase of the search.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
21
Q

Which components do the particle consists of and which does the swarm have in addition?

A

solution candidate for the position
velocity
local best solution = local attractor
swarm = particles (1,..,N), global best position = global attractor

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
22
Q

Show the components of a particle and the swarm graphically (without parameters). Include the next position of the particle additionally

A
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
23
Q

Write down the Movement equations with all parameters and variables. Name the dimensionallities and what kind of parameters they are.

A
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
24
Q

Name 3 different kind of search querys. Which of these is used in HITS?

A

specific query
broad-topic query
similar-page query

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
25
Q

What are authorities and hubs? How do they relate to emergence?

A

Authority: the (most) relevant pages to the query
Hub: Those pages that know the authorities (authorities for authorities)
metrics can be used to measure the emergent property of a web page to provide good (best?) information regarding the query

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
26
Q

What does HITS stand for?

A

Hyperlink-Induced Topic Search, or Hypertext-Induced Topic Search

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
27
Q

What three properties must the base graph S possess, so that HITS works best?

A
  • relatively small
  • contains many relevant pages
  • contains many pages with the highest authority
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
28
Q

What does x and y stands for in the HITS-algorithm

A

authority weight x_p and the hub weight y_p
The larger the value of x_p, the more suitable the page p is as an answer to the query
The larger the value of y_p, the more “significant” the page is as a hub.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
29
Q

What is the formula for updating x and y (write a→b for a page pointing from a to b) in the HITS-alg.?

A

The new authority is the sum of the hub of its parents
The new hub is the sum of the authority of its children

30
Q

What are the 2 mathematical useful properties of A^TA and AA^T -matrices?

A

is symmetric
all eigenvalues are real-valued → the principal eigenvector has only real entries

31
Q

What two additional heuristics can we apply to the base graph S in order to improve the quality of the search result?

A

-Delete internal (w.r.t domain name) links, keep external links → eliminate navigation links
-Allow only m (4 through 8) links from a domain to a page (Avoiding “This page was created using …”)

32
Q

How can the adjacency matrix of the graph be used for the calculation of x and y in HITS algorithm?

A
33
Q

What is ACO?

A

ACO = Ant Colony Optimization
metaheuristic for solving (usually hard) combinational optimization problems
ant inspired algorithm

34
Q

How does ACO work?

A

-Exchange information using (artificial) pheromones
-Randomized search heuristic is implemented for constructing candidate solutions → make probabilistic decisions depending on artificial pheromone intensities and other heuristic information that may be available.
-Heuristic information in the ant algorithm i.e. the pheromone trails, is adjusted during the search to reflect the progress of the search.

35
Q

Describe 4 components of the ACO problem

A

D: the set of problem instances, a.k.a input
S(I) for I in D: set of admissible solutions w.r.t. Input I
The objective function f
Goal in {min, max}

36
Q

What are the algorithm steps of ACO?

A
37
Q

Give the most widely used rule for determining the probability of choosing the next path in ACO

A
38
Q

Give the formula describing how pheromones are updated in ACO

A
39
Q

Which heuristic information can be used for TSP using ACO?

A

reciprocal of distance points i and j

40
Q

Which emission functions can be used for TSP using ACO?

A

Ant Density, Ant Quantity, Ant Cycle

41
Q

How is the pheromone values updated in ACO?

A

The emission of new pheromones: this involves selecting the best of the candidate solutions found so far, and “rewarding” their solution extensions with pheromones.
The (partial) “evaporation” of old pheromones: this is necessary to avoid early convergence towards suboptimal solutions. As a result, the ants “forget” a more distant past, which leads to searching new regions of the search space as well.

42
Q

Draw the evolutionary cycle. What do the parameters mean?

A
43
Q

What is the evolutionary algorithm about?

A

meta-heuristic that uses observations from the theory of evolution to solve optimization problems of various types.

44
Q

What do reproduction and selection in EA mean?

A

Reproduction involves
inheritance
variation and mutation
Selection:
only the best-fitting individuals of the parent generation should be selected for crossover (recombination) (parent selection),
only the best individuals of parents and children should be selected for the next generation (survivor selection)

45
Q

What is the difference between genotype and phenotype in ACO?

A

“genotype” refers to the coding of an individual,
“phenotype” refers to the coded individual itself.

46
Q

What is elitism?

A

the best individual is automatically carried over to the next generation during the selection

47
Q

Name and explain possible selection operators.

A

Tournament Selection:
Two or more individuals are randomly drawn from the population, with the best individual surviving or being allowed to recombine.
Rank Selection:
a ranking list of the N individuals is created with respect to their quality –>The individual Ik will then be selected with probability

48
Q

Name and explain possible Crossover Operators.

A

-One-point crossover
A crossover point is chosen at random.
For each parent individual, the first part from the beginning through the crossover point is kept,
the remaining numbers are given the order they have in the other parent.
-Partially matched crossover (PMX)
Randomly select two crossover points XP1 and XP2, and interchange the intermediate partial permutation. Now fill up the remaining positions with numbers from the corresponding parent. If this would result in a duplicate number x, consider the partial permutation between the crossover points as a mapping of x to the number in the other parent, and use the assigned number

49
Q

Name and explain possible Mutation Operators.

A

-Swap mutation operator:
Two numbers of the permutation are randomly chosen and then interchanged.
-Inversion mutation operator:
Two numbers of the permutation are randomly chosen and then the intermediate sequence (including the chosen numbers) is reversed.

50
Q

Name and explain the objective functions that measure the sortedness of a permutation in (1,1)-EA for sorting

A

INV: number of pairs in the correct order
HAM: number of items already at the final position.
LAS: length of a longest ascending subsequence

51
Q

What is a P2P network?

A

communication network btw computers on the internet
In which there is no central control
And no reliable partners

52
Q

How does Napster work?

A

Client-server structure
Server maintains
Index with metadata e.g. file name, date, etc.
Table of connections of the participating clients
Table of all files of the participating clients
→ File indexing and centralized index server
Query/Reply
A client asks (query) for file names
Server searches for matching participants
Server replies (reply) who owns the file
Request client download from file-owning client → Peer-to-Peer Downloading → File transfer

53
Q

Draw the principle of Napster. Your sketch should necessarily include a file.

A
54
Q

Name one advantage and disadvantage of Napster each.

A

(+) Simple
Files are found quickly and efficiently
(-)The centralized structure facilitates censorship, hostile interference, and technical glitches e.g denial of service attack
Does not scale due to e.g. increasing nr. participants, finite memory on the server

55
Q

What was the novelty of Napster?

A

File sharing
Peer to Peer Downloading
Otherwise, Napster is not an acceptable P2P network solution

56
Q

How does Gnutella work?

A

File sharing system works without central structure, using neighborhood Lists
File request
Is sent to all neighbors
They send them to their neighbors
Up to a given number of hops
TTL field
Protocol
Query: request for the file is forwarded up to the number of TTL hops
Query-hits: reverse path response
If a file was found, direct download

57
Q

Draw the principle of Gnutella. In your sketch, there should definitely be a file.

A
58
Q

What does TTL mean?

A

Maximum number of hops a query can travel from one node to another before it expires or is discarded

59
Q

What is the advantage of Gnutella?

A

Distributed network structure
Scalable network

60
Q

What is the main disadvantage of Gnutella?

A
61
Q

How to avoid the main disadvantage of Gnutella?

A

-By TTL an implicit network partitioning takes place for queries i.e. the query becomes limited to a specific local region of the network
-Inquiry success is low due to Long paths, high latency

62
Q

What is a hash table as P2P network?

A

each peer represents a memory location
a hash function known to all peers
peers are connected as a chain
Search happens by calculating the hash function and going to peer with the address of the calculated result of the function along the line.

63
Q

What is the advantage and disadvantage of DHT?

A

Distributed Hash Table
Peers are hashed to a location and assigned ranges of the hash function’s value range
Data is also hashed → assigned peers depending on the area

64
Q

What is CAN?

A

CAN= Content Addressable Network
Data is mapped into the square by the hash function (two-valued or d-valued hash function)

65
Q

Draw the connection structures between the mapped peers of CAN.

A
66
Q

What is the formula for the expected value E[A(p)] when n is the number of peers?

A
67
Q

What is the probability that three peers do not meet in the mapped area?

A
68
Q

What is the potential risk of the increasing complexity of technical systems?

A

More difficult to control
Hard to predict error or failure

69
Q
A
69
Q
A

Swarm Intelligence (1 decks)

Brainscape helps you reach your goals faster, through stronger study habits.
© 2024 Bold Learning Solutions. Terms and Conditions