SlideShare a Scribd company logo
1 of 39
Download to read offline
Why Advanced Population Initialization Techniques
Perform Poorly in High Dimension?
Borhan Kazimipour
Xiaodong Li
A.K. Qin
Outlines
1. Introduction
2. Background
3. Questions
4. Experiments
5. Results
6. Conclusion
SEAL 2014, Dunedin, NZ 2Why Advanced PITs Perform Poorly in HD?
Outlines
1. Introduction
2. Background
3. Questions
4. Experiments
5. Results
6. Conclusion
SEAL 2014, Dunedin, NZ 3Why Advanced PITs Perform Poorly in HD?
Definition of Population Initialization
• Definition:
– Initialization is the task of generating a set of initial points as potential solutions of
an optimization problem. These values are seen as the first position (or distribution)
of the individuals in the first generation.
• Common Parameters:
– Population size
– Number of variables or dimensionality (given)
– Variables range (given)
• Note: In this study our main focus is on continuous techniques capable of generating
real-value numbers in continuous spaces.
SEAL 2014, Dunedin, NZ 4Why Advanced PITs Perform Poorly in HD?
Importance of Population Initialization
• Why studying population initialization is important?
– Popularity: All population-based algorithms, including EA, need a population
initialization module.
– “initialize population randomly” is the most widely used expression in EA community!
– Variety: Lots of different population initialization techniques are proposed, so far.
– About 100 population initialization techniques are proposed so far*.
– Effectiveness: Clearly, starting from a good position makes it easier and faster to
achieve the aim, than starting from a bad one.
– “Advanced initialization techniques can increase the probability of finding global optima,
reduce the variation of the final results, decrease the computational costs and improve the
solution(s) quality.” *
– Inconsistency(!): Some controversy findings have been reported.
– “For example, one claimed that the desirable effect of uniformity of initial population is more
significant in high dimensions (up to 50 dimensions) while another study, in contrast,
claimed that uniform initialization techniques loose their effectiveness in problems of 12 or
more dimensions.” *
* B. Kazimipour, X. Li, and A. K. Qin. "A review of population initialization techniques for evolutionary algorithms." In Evolutionary Computation (CEC), 2014 IEEE Congress on,
pp. 2585-2592. IEEE, 2014.
SEAL 2014, Dunedin, NZ 5Why Advanced PITs Perform Poorly in HD?
Outlines
1. Introduction
2. Background
3. Questions
4. Experiments
5. Results
6. Conclusion
SEAL 2014, Dunedin, NZ 6Why Advanced PITs Perform Poorly in HD?
Definitions of Randomness
• True Random:
– A true random sequence is usually described as a sequence having strong
properties such as complete unpredictability, incompressibility and irregularity.
– Some believe true random sequences do not exist (theoretical drawback).
– There is no tool to proof a given sequence is truly random (empirical drawback).
• Computational Random:
– A sequence is computationally random if it passes some tests on the properties of
true randomness e.g., unpredictability, and incompressibility.
• Statistical Random:
– A sequence is statistically random if it passes some tests on the statistical
(distributional) properties of true random sequences e.g., uniformity.
SEAL 2014, Dunedin, NZ 7Why Advanced PITs Perform Poorly in HD?
Continuum of RandomnessCompletely
Deterministic
Truly
Random
• In this work, we follow the technique proposed in [1] to categorise PITs based on
randomness:
Does output
depend on
initial seed?
Stochastic Deterministic
YES NO
Measuring Randomness
SEAL 2014, Dunedin, NZ 8Why Advanced PITs Perform Poorly in HD?
Categorization based on Randomness
SEAL 2014, Dunedin, NZ 9Why Advanced PITs Perform Poorly in HD?
Population Initialization Techniques
Stochastic
Pseudo-Random
Number Generator
Chaotic Number
Generator
Deterministic
Quasi-Random
Sequence
Uniform Experimental
Design
Stochastic vs. Deterministic
Stochastic
• Definition:
– Their results depend on initial seeds.
• Properties:
– Unpredictable (computationally)
– Irregularity
• Examples:
– Pseudo-Random Number Generator
(PRNG)
– e.g. WELL, KISS, and Mersenne Twister
– Chaotic Number Generator (CNG)
– e.g. Tent, Logistic and Sine
Deterministic
• Definition:
– They always generate the same
population regardless of any initial
seed.
• Properties:
– Population uniformity is more
important than randomness or
unpredictability.
• Examples:
– Quasi-random Sequence
– e.g. Sobol, Halton
– Uniform Experimental Design
– e.g. Latin hypercube, good lattice points and
orthogonal design
SEAL 2014, Dunedin, NZ Why Advanced PITs Perform Poorly in HD? 10
Outlines
1. Introduction
2. Background
3. Questions
4. Experiments
5. Results
6. Conclusion
SEAL 2014, Dunedin, NZ 11Why Advanced PITs Perform Poorly in HD?
Questions
Goal
• Research Question:
– Why EAs do not receive great benefit from advanced population initialization
techniques when dimensionality of problems are very high?
• Hypothesis:
– The uniformity of population for both simple and advanced techniques drop to the
same level when dimensionality grows.
SEAL 2014, Dunedin, NZ 12Why Advanced PITs Perform Poorly in HD?
Questions…
Two experiments
Part A
(baseline technique)
• Goal: Study the trend of population
uniformity when generated by popular
but simple techniques*.
• Research Questions:
1. How much the uniformity of a
population can be affected by
dimensionality?
2. Is it possible to enhance the
uniformity of initial population in
high dimensional spaces by
increasing the population size?
Part B
(advanced techniques)
• Goal: Compare the performance of
advanced initialization techniques with a
commonly used technique*.
• Research Questions:
1. Can adopting advanced
initialization techniques significantly
improve population uniformity?
2. How population size affects
performance of advanced
initializers?
SEAL 2014, Dunedin, NZ Why Advanced PITs Perform Poorly in HD? 13
*Random number generators (RNG) are the most widely used initializers in the field of EA.
Questions…
Quality measures
In both parts, we use discrepancy values to measure quality of populations.
• Definition of discrepancy:
– Literally, discrepancy means non-uniformity.
– Technically, discrepancy measures are tools for determining non-uniformity level of
a given point set.
– Point sets with low discrepancy are those with high level of uniformity.
• Variations of discrepancy:
– Star L2-discrepancy
– Centred L2-discrepancy*
– Modified L2-discrepancy
– Symmetric L2-discrepancy
– Wrap-around L2-discrepancy
* Centred L2-discrepancy (CD) is used in this study.
SEAL 2014, Dunedin, NZ 14Why Advanced PITs Perform Poorly in HD?
Questions…
Analytic formulas
• L2-discrepancy (D: dimensionality, N: population size, P: population, xi,j: ith value of jth
individual)
• Centred L2-discrepancy (D: dimensionality, N: population size, P: population, xi,j: ith
value of jth individual)
SEAL 2014, Dunedin, NZ 15Why Advanced PITs Perform Poorly in HD?
Questions…
Why we chose discrepancy?
Discrepancy measures with analytic formulas are used in this study because:
Discrepancy values are not affected by the features of benchmarked problems,
employed EAs or their parameters.
– Unlike final fitness value and success rate.
Discrepancy measures can be easily applied to all kinds of real-value populations.
– Unlike DieHard and TestU01 which can only be applied on stochastic population.
Discrepancy measures having analytic formulas are faster than similar
iterative/recursive algorithms (ideal for large and high dimensional populations).
– Unlike early variants of Lp-discrepancy.
SEAL 2014, Dunedin, NZ 16Why Advanced PITs Perform Poorly in HD?
Outlines
1. Introduction
2. Background
3. Questions
4. Experiments
5. Results
6. Conclusion
SEAL 2014, Dunedin, NZ 17Why Advanced PITs Perform Poorly in HD?
Experiments
Setup
• Six population initialization techniques are selected to study.
• Three stochastic and three deterministic techniques are included in the experiments.
• RNG, which is the most common and simple initializer is chosen as the control method.
SEAL 2014, Dunedin, NZ 18Why Advanced PITs Perform Poorly in HD?
Experiments…
Setup
• In both parts:
– 20 different dimension sizes are examined (2 ≤ D ≤ 1,000).
– 20 different population sizes are examined (10 ≤ N ≤ 10,000).
– Each experiment is run for 25 times:
– 25 unique initial seeds are used for stochastic techniques
– 25 unique sequences are used for deterministic techniques (skip schema)
• Part A (baseline technique)
– Only performance of RNG is examined in different situations.
• Part B (advanced technique)
– Performance of advanced techniques are compared with the baseline (RNG).
SEAL 2014, Dunedin, NZ 19Why Advanced PITs Perform Poorly in HD?
Outlines
1. Introduction
2. Background
3. Questions
4. Experiments
5. Results
– Part A
– Part B
6. Conclusion
SEAL 2014, Dunedin, NZ 20Why Advanced PITs Perform Poorly in HD?
Results
Part A – Dimensionality effect
SEAL 2014, Dunedin, NZ 21Why Advanced PITs Perform Poorly in HD?
Results
Part A – Dimensionality effect
SEAL 2014, Dunedin, NZ 22Why Advanced PITs Perform Poorly in HD?
• Discrepancy grows (i.e., uniformity drops) exponentially when the dimensionality increases.
– Discrepancy of 10,000 points in 50 dimensions is comparable with the discrepancy of 10 points
in 30 dimensions!
– 66% growth in dimensionality demands 100,000% increase in population size to recover the
uniformity!
• For D ≤ 50, a large population size may lessen the undesirable effect of dimensionality (zoomed in
the graph)
Results
Part A – Low dimensions
SEAL 2014, Dunedin, NZ 23Why Advanced PITs Perform Poorly in HD?
Results
Part A – Low dimensions
SEAL 2014, Dunedin, NZ 24Why Advanced PITs Perform Poorly in HD?
• Population size has no considerable effect on the uniformity of very small-sized problems (D ≤ 10).
• For 30 ≤ D ≤ 50, population size has a significant effect on uniformity such that it can be improved
10 to 20 times in the CD scale.
• The magnitude of improvements falls rapidly such that increasing population size beyond 1,000
points shows only a minimal improvement.
Results
Part A – Medium dimensions
SEAL 2014, Dunedin, NZ 25Why Advanced PITs Perform Poorly in HD?
Results
Part A – Medium dimensions
SEAL 2014, Dunedin, NZ 26Why Advanced PITs Perform Poorly in HD?
• Increase in population size significantly lessens the effect of dimensionality (specially N ≤ 200)
• The magnitudes of improvements fall as population grows.
Results
Part A – High dimensions
SEAL 2014, Dunedin, NZ 27Why Advanced PITs Perform Poorly in HD?
Results
Part A – High dimensions
SEAL 2014, Dunedin, NZ 28Why Advanced PITs Perform Poorly in HD?
• Uniformity of populations in spaces of above 100 dimensions is so weak that increasing population
size from 1,000 to 10,000 cannot recover it.
• The feasible and reasonable population size for very large-scale problems (100 ≤ D) is surprisingly
less than 1,000 points.
• It does not imply N has no effect in D > 100. Instead, it means N must be astronomically large to
achieve a significant enhancement. Since evaluating high dimensional populations in that magnitude
is currently computationally infeasible, keeping it around 1,000 points is more practical and
reasonable.
Outlines
1. Introduction
2. Background
3. Questions
4. Experiments
5. Results
– Part A
– Part B
6. Conclusion
SEAL 2014, Dunedin, NZ 29Why Advanced PITs Perform Poorly in HD?
Results
Part B – Improvement
Improvement over common technique:
• To compare advanced initialization techniques with a common RNG, we propose a
simple formula reflecting relative improvement achieved from each advanced
technique:
where Pc is the population generated by the control technique (RNG), and Pi is the
population produced by the ith advanced initialization technique and CD is centred L2-
discrepancy.
SEAL 2014, Dunedin, NZ 30Why Advanced PITs Perform Poorly in HD?
Results
Part B – Low dimensions
SEAL 2014, Dunedin, NZ 31Why Advanced PITs Perform Poorly in HD?
Results
Part B – Low dimensions
SEAL 2014, Dunedin, NZ 32Why Advanced PITs Perform Poorly in HD?
• Some techniques (TNT and SBL) are successful in improving the common initializer (RNG), although
the biggest improvement in 2 ≤ D ≤ 50 is less than 20%.
• Some techniques (GLP) are very sensitive to population size, others (SBL) are more stable.
• For D ≤ 50 ,with no exception, all techniques work relatively better when population size increases.
• Mixed good and bad results can be expected from both categories of initialization techniques*.
*B. Kazimipour, X. Li, and A. K. Qin. "Initialization methods for large scale global optimization." In Evolutionary Computation (CEC), 2013 IEEE Congress on, pp. 2750-2757. IEEE, 2013.
Results
Part B – Medium and High dimensions
SEAL 2014, Dunedin, NZ 33Why Advanced PITs Perform Poorly in HD?
Results
Part B – Medium and High dimensions
SEAL 2014, Dunedin, NZ 34Why Advanced PITs Perform Poorly in HD?
• All trends converge to one of the three values: 0%, -25% and -80%.
• This clearly shows that employing advanced initialization techniques provides no significant
improvement in high dimensions, at least in terms of uniformity.
Results
Part B – Medium and High dimensions
SEAL 2014, Dunedin, NZ 35Why Advanced PITs Perform Poorly in HD?
• Even increasing population size from 10 to 10,000 does not result in any relative improvement
• SBL with 10 and TNT with all population sizes perform almost the same as RNG.
• The others, however, perform poorly in comparison with a RNG having the same population size*.
* B. Kazimipour, X. Li, and A. K. Qin. "Effects of population initialization on differential evolution for large scale optimization." In Evolutionary Computation (CEC), 2014 IEEE
Congress on, pp. 2404-2411. IEEE, 2014.
Outlines
1. Introduction
2. Background
3. Questions
4. Experiments
5. Results
6. Conclusion
SEAL 2014, Dunedin, NZ 36Why Advanced PITs Perform Poorly in HD?
Conclusion
What we did
• We investigate the reasons that causes advanced population initialization techniques to
perform as poor as simple RNG in high dimensional spaces.
• We also studied the effect of population size on the quality (uniformity) of the resulting
populations.
• We studied:
– 6 techniques (3 deterministic and 3 stochastic),
– 20 dimension sizes (up to 1,000),
– 20 population sizes (up to 10,000),
– thorough 25 runs.
SEAL 2014, Dunedin, NZ 37Why Advanced PITs Perform Poorly in HD?
Conclusion
What we observed
• The uniformity of initial population drops exponentially when dimensionality rises
linearly.
• Increasing population size up to a computationally feasible bound cannot maintain
uniformity (except for some small and medium-sized spaces).
• The advanced initializers are as vulnerable to the curse of dimensionality as simple
RNG.
• Adopting advanced initializers in medium and large-scale spaces does not result in
any significant improvement.
• Some advanced techniques are even more sensitive to the adverse effect of
dimensionality than the simple RNG.
SEAL 2014, Dunedin, NZ 38Why Advanced PITs Perform Poorly in HD?
We only recommend the use of advanced techniques when the population and
dimension sizes are small. In higher dimensional spaces or when the population size
is relatively large, no significant improvement is excepted from advanced techniques.
Thank you
☺☺☺☺
Any question or comment?
39SEAL 2014, Dunedin, NZ Why Advanced PITs Perform Poorly in HD?

More Related Content

Similar to Why advanced population initialization techniques perform poorly in high dimension

Lecture_4_Data_Gathering_and_Analysis.pdf
Lecture_4_Data_Gathering_and_Analysis.pdfLecture_4_Data_Gathering_and_Analysis.pdf
Lecture_4_Data_Gathering_and_Analysis.pdfAbdullahOmar64
 
IITSEC Presentation on Learning in Virtual Worlds
IITSEC Presentation on Learning in Virtual WorldsIITSEC Presentation on Learning in Virtual Worlds
IITSEC Presentation on Learning in Virtual Worldstaoirene
 
MLSEV Virtual. State of the Art in ML
MLSEV Virtual. State of the Art in MLMLSEV Virtual. State of the Art in ML
MLSEV Virtual. State of the Art in MLBigML, Inc
 
Machine Learning Foundations
Machine Learning FoundationsMachine Learning Foundations
Machine Learning FoundationsAlbert Y. C. Chen
 
Learning analytics and accessibility – #calrg 2015
Learning analytics and accessibility – #calrg 2015Learning analytics and accessibility – #calrg 2015
Learning analytics and accessibility – #calrg 2015Martyn Cooper
 
Mmig talk jan 245 2011
Mmig talk jan 245 2011Mmig talk jan 245 2011
Mmig talk jan 245 2011Brock Dubbels
 
Replicable Evaluation of Recommender Systems
Replicable Evaluation of Recommender SystemsReplicable Evaluation of Recommender Systems
Replicable Evaluation of Recommender SystemsAlejandro Bellogin
 
Adjusting for Differential Item Functioning in the EQ-5D-5L Using Externally-...
Adjusting for Differential Item Functioning in the EQ-5D-5L Using Externally-...Adjusting for Differential Item Functioning in the EQ-5D-5L Using Externally-...
Adjusting for Differential Item Functioning in the EQ-5D-5L Using Externally-...Office of Health Economics
 
Meta analysis presentation-sim vs. no sim
Meta analysis presentation-sim vs. no simMeta analysis presentation-sim vs. no sim
Meta analysis presentation-sim vs. no simJ'ette Novakovich
 
Large Scale GAN Training for High Fidelity Natural Image Synthesis
Large Scale GAN Training for High Fidelity Natural Image SynthesisLarge Scale GAN Training for High Fidelity Natural Image Synthesis
Large Scale GAN Training for High Fidelity Natural Image SynthesisSeunghyun Hwang
 
Bridging the Gap: Machine Learning for Ubiquitous Computing -- Evaluation
Bridging the Gap: Machine Learning for Ubiquitous Computing -- EvaluationBridging the Gap: Machine Learning for Ubiquitous Computing -- Evaluation
Bridging the Gap: Machine Learning for Ubiquitous Computing -- EvaluationThomas Ploetz
 
Impact Evaluation: Balancing Rigor with Reality
Impact Evaluation: Balancing Rigor with RealityImpact Evaluation: Balancing Rigor with Reality
Impact Evaluation: Balancing Rigor with RealityDonna Smith-Moncrieffe
 
Deep Q-learning from Demonstrations DQfD
Deep Q-learning from Demonstrations DQfDDeep Q-learning from Demonstrations DQfD
Deep Q-learning from Demonstrations DQfDAmmar Rashed
 
Aligning tests to standards
Aligning tests to standardsAligning tests to standards
Aligning tests to standardsFariba Chamani
 
How Data Scientists Make Reliable Decisions with Data
How Data Scientists Make Reliable Decisions with DataHow Data Scientists Make Reliable Decisions with Data
How Data Scientists Make Reliable Decisions with DataTa-Wei (David) Huang
 
Mixed Effects Models - Random Intercepts
Mixed Effects Models - Random InterceptsMixed Effects Models - Random Intercepts
Mixed Effects Models - Random InterceptsScott Fraundorf
 
Peeking behind the test: insights and innovations from the Medical Council of...
Peeking behind the test: insights and innovations from the Medical Council of...Peeking behind the test: insights and innovations from the Medical Council of...
Peeking behind the test: insights and innovations from the Medical Council of...MedCouncilCan
 
eMOOCs2015 Does peer grading work?
eMOOCs2015 Does peer grading work?eMOOCs2015 Does peer grading work?
eMOOCs2015 Does peer grading work?Rémi Bachelet
 

Similar to Why advanced population initialization techniques perform poorly in high dimension (20)

Lecture_4_Data_Gathering_and_Analysis.pdf
Lecture_4_Data_Gathering_and_Analysis.pdfLecture_4_Data_Gathering_and_Analysis.pdf
Lecture_4_Data_Gathering_and_Analysis.pdf
 
IITSEC Presentation on Learning in Virtual Worlds
IITSEC Presentation on Learning in Virtual WorldsIITSEC Presentation on Learning in Virtual Worlds
IITSEC Presentation on Learning in Virtual Worlds
 
MLSEV Virtual. State of the Art in ML
MLSEV Virtual. State of the Art in MLMLSEV Virtual. State of the Art in ML
MLSEV Virtual. State of the Art in ML
 
Machine Learning Foundations
Machine Learning FoundationsMachine Learning Foundations
Machine Learning Foundations
 
Learning analytics and accessibility – #calrg 2015
Learning analytics and accessibility – #calrg 2015Learning analytics and accessibility – #calrg 2015
Learning analytics and accessibility – #calrg 2015
 
Mmig talk jan 245 2011
Mmig talk jan 245 2011Mmig talk jan 245 2011
Mmig talk jan 245 2011
 
Replicable Evaluation of Recommender Systems
Replicable Evaluation of Recommender SystemsReplicable Evaluation of Recommender Systems
Replicable Evaluation of Recommender Systems
 
Adjusting for Differential Item Functioning in the EQ-5D-5L Using Externally-...
Adjusting for Differential Item Functioning in the EQ-5D-5L Using Externally-...Adjusting for Differential Item Functioning in the EQ-5D-5L Using Externally-...
Adjusting for Differential Item Functioning in the EQ-5D-5L Using Externally-...
 
Meta analysis presentation-sim vs. no sim
Meta analysis presentation-sim vs. no simMeta analysis presentation-sim vs. no sim
Meta analysis presentation-sim vs. no sim
 
Large Scale GAN Training for High Fidelity Natural Image Synthesis
Large Scale GAN Training for High Fidelity Natural Image SynthesisLarge Scale GAN Training for High Fidelity Natural Image Synthesis
Large Scale GAN Training for High Fidelity Natural Image Synthesis
 
Cemteamie4
Cemteamie4Cemteamie4
Cemteamie4
 
Bridging the Gap: Machine Learning for Ubiquitous Computing -- Evaluation
Bridging the Gap: Machine Learning for Ubiquitous Computing -- EvaluationBridging the Gap: Machine Learning for Ubiquitous Computing -- Evaluation
Bridging the Gap: Machine Learning for Ubiquitous Computing -- Evaluation
 
05 Programme evaluation
05 Programme evaluation05 Programme evaluation
05 Programme evaluation
 
Impact Evaluation: Balancing Rigor with Reality
Impact Evaluation: Balancing Rigor with RealityImpact Evaluation: Balancing Rigor with Reality
Impact Evaluation: Balancing Rigor with Reality
 
Deep Q-learning from Demonstrations DQfD
Deep Q-learning from Demonstrations DQfDDeep Q-learning from Demonstrations DQfD
Deep Q-learning from Demonstrations DQfD
 
Aligning tests to standards
Aligning tests to standardsAligning tests to standards
Aligning tests to standards
 
How Data Scientists Make Reliable Decisions with Data
How Data Scientists Make Reliable Decisions with DataHow Data Scientists Make Reliable Decisions with Data
How Data Scientists Make Reliable Decisions with Data
 
Mixed Effects Models - Random Intercepts
Mixed Effects Models - Random InterceptsMixed Effects Models - Random Intercepts
Mixed Effects Models - Random Intercepts
 
Peeking behind the test: insights and innovations from the Medical Council of...
Peeking behind the test: insights and innovations from the Medical Council of...Peeking behind the test: insights and innovations from the Medical Council of...
Peeking behind the test: insights and innovations from the Medical Council of...
 
eMOOCs2015 Does peer grading work?
eMOOCs2015 Does peer grading work?eMOOCs2015 Does peer grading work?
eMOOCs2015 Does peer grading work?
 

Recently uploaded

Dubai Call Girls Peeing O525547819 Call Girls Dubai
Dubai Call Girls Peeing O525547819 Call Girls DubaiDubai Call Girls Peeing O525547819 Call Girls Dubai
Dubai Call Girls Peeing O525547819 Call Girls Dubaikojalkojal131
 
Sealdah % High Class Call Girls Kolkata - 450+ Call Girl Cash Payment 8005736...
Sealdah % High Class Call Girls Kolkata - 450+ Call Girl Cash Payment 8005736...Sealdah % High Class Call Girls Kolkata - 450+ Call Girl Cash Payment 8005736...
Sealdah % High Class Call Girls Kolkata - 450+ Call Girl Cash Payment 8005736...HyderabadDolls
 
Predicting HDB Resale Prices - Conducting Linear Regression Analysis With Orange
Predicting HDB Resale Prices - Conducting Linear Regression Analysis With OrangePredicting HDB Resale Prices - Conducting Linear Regression Analysis With Orange
Predicting HDB Resale Prices - Conducting Linear Regression Analysis With OrangeThinkInnovation
 
Top profile Call Girls In Satna [ 7014168258 ] Call Me For Genuine Models We ...
Top profile Call Girls In Satna [ 7014168258 ] Call Me For Genuine Models We ...Top profile Call Girls In Satna [ 7014168258 ] Call Me For Genuine Models We ...
Top profile Call Girls In Satna [ 7014168258 ] Call Me For Genuine Models We ...nirzagarg
 
SAC 25 Final National, Regional & Local Angel Group Investing Insights 2024 0...
SAC 25 Final National, Regional & Local Angel Group Investing Insights 2024 0...SAC 25 Final National, Regional & Local Angel Group Investing Insights 2024 0...
SAC 25 Final National, Regional & Local Angel Group Investing Insights 2024 0...Elaine Werffeli
 
Aspirational Block Program Block Syaldey District - Almora
Aspirational Block Program Block Syaldey District - AlmoraAspirational Block Program Block Syaldey District - Almora
Aspirational Block Program Block Syaldey District - AlmoraGovindSinghDasila
 
Lake Town / Independent Kolkata Call Girls Phone No 8005736733 Elite Escort S...
Lake Town / Independent Kolkata Call Girls Phone No 8005736733 Elite Escort S...Lake Town / Independent Kolkata Call Girls Phone No 8005736733 Elite Escort S...
Lake Town / Independent Kolkata Call Girls Phone No 8005736733 Elite Escort S...HyderabadDolls
 
Fun all Day Call Girls in Jaipur 9332606886 High Profile Call Girls You Ca...
Fun all Day Call Girls in Jaipur   9332606886  High Profile Call Girls You Ca...Fun all Day Call Girls in Jaipur   9332606886  High Profile Call Girls You Ca...
Fun all Day Call Girls in Jaipur 9332606886 High Profile Call Girls You Ca...kumargunjan9515
 
Digital Transformation Playbook by Graham Ware
Digital Transformation Playbook by Graham WareDigital Transformation Playbook by Graham Ware
Digital Transformation Playbook by Graham WareGraham Ware
 
Diamond Harbour \ Russian Call Girls Kolkata | Book 8005736733 Extreme Naught...
Diamond Harbour \ Russian Call Girls Kolkata | Book 8005736733 Extreme Naught...Diamond Harbour \ Russian Call Girls Kolkata | Book 8005736733 Extreme Naught...
Diamond Harbour \ Russian Call Girls Kolkata | Book 8005736733 Extreme Naught...HyderabadDolls
 
Case Study 4 Where the cry of rebellion happen?
Case Study 4 Where the cry of rebellion happen?Case Study 4 Where the cry of rebellion happen?
Case Study 4 Where the cry of rebellion happen?RemarkSemacio
 
Top profile Call Girls In Hapur [ 7014168258 ] Call Me For Genuine Models We ...
Top profile Call Girls In Hapur [ 7014168258 ] Call Me For Genuine Models We ...Top profile Call Girls In Hapur [ 7014168258 ] Call Me For Genuine Models We ...
Top profile Call Girls In Hapur [ 7014168258 ] Call Me For Genuine Models We ...nirzagarg
 
Charbagh + Female Escorts Service in Lucknow | Starting ₹,5K To @25k with A/C...
Charbagh + Female Escorts Service in Lucknow | Starting ₹,5K To @25k with A/C...Charbagh + Female Escorts Service in Lucknow | Starting ₹,5K To @25k with A/C...
Charbagh + Female Escorts Service in Lucknow | Starting ₹,5K To @25k with A/C...HyderabadDolls
 
Introduction to Statistics Presentation.pptx
Introduction to Statistics Presentation.pptxIntroduction to Statistics Presentation.pptx
Introduction to Statistics Presentation.pptxAniqa Zai
 
Jual obat aborsi Bandung ( 085657271886 ) Cytote pil telat bulan penggugur ka...
Jual obat aborsi Bandung ( 085657271886 ) Cytote pil telat bulan penggugur ka...Jual obat aborsi Bandung ( 085657271886 ) Cytote pil telat bulan penggugur ka...
Jual obat aborsi Bandung ( 085657271886 ) Cytote pil telat bulan penggugur ka...Klinik kandungan
 
20240412-SmartCityIndex-2024-Full-Report.pdf
20240412-SmartCityIndex-2024-Full-Report.pdf20240412-SmartCityIndex-2024-Full-Report.pdf
20240412-SmartCityIndex-2024-Full-Report.pdfkhraisr
 
Top profile Call Girls In Rohtak [ 7014168258 ] Call Me For Genuine Models We...
Top profile Call Girls In Rohtak [ 7014168258 ] Call Me For Genuine Models We...Top profile Call Girls In Rohtak [ 7014168258 ] Call Me For Genuine Models We...
Top profile Call Girls In Rohtak [ 7014168258 ] Call Me For Genuine Models We...nirzagarg
 
Top profile Call Girls In Bihar Sharif [ 7014168258 ] Call Me For Genuine Mod...
Top profile Call Girls In Bihar Sharif [ 7014168258 ] Call Me For Genuine Mod...Top profile Call Girls In Bihar Sharif [ 7014168258 ] Call Me For Genuine Mod...
Top profile Call Girls In Bihar Sharif [ 7014168258 ] Call Me For Genuine Mod...nirzagarg
 
RESEARCH-FINAL-DEFENSE-PPT-TEMPLATE.pptx
RESEARCH-FINAL-DEFENSE-PPT-TEMPLATE.pptxRESEARCH-FINAL-DEFENSE-PPT-TEMPLATE.pptx
RESEARCH-FINAL-DEFENSE-PPT-TEMPLATE.pptxronsairoathenadugay
 
Top profile Call Girls In dimapur [ 7014168258 ] Call Me For Genuine Models W...
Top profile Call Girls In dimapur [ 7014168258 ] Call Me For Genuine Models W...Top profile Call Girls In dimapur [ 7014168258 ] Call Me For Genuine Models W...
Top profile Call Girls In dimapur [ 7014168258 ] Call Me For Genuine Models W...gajnagarg
 

Recently uploaded (20)

Dubai Call Girls Peeing O525547819 Call Girls Dubai
Dubai Call Girls Peeing O525547819 Call Girls DubaiDubai Call Girls Peeing O525547819 Call Girls Dubai
Dubai Call Girls Peeing O525547819 Call Girls Dubai
 
Sealdah % High Class Call Girls Kolkata - 450+ Call Girl Cash Payment 8005736...
Sealdah % High Class Call Girls Kolkata - 450+ Call Girl Cash Payment 8005736...Sealdah % High Class Call Girls Kolkata - 450+ Call Girl Cash Payment 8005736...
Sealdah % High Class Call Girls Kolkata - 450+ Call Girl Cash Payment 8005736...
 
Predicting HDB Resale Prices - Conducting Linear Regression Analysis With Orange
Predicting HDB Resale Prices - Conducting Linear Regression Analysis With OrangePredicting HDB Resale Prices - Conducting Linear Regression Analysis With Orange
Predicting HDB Resale Prices - Conducting Linear Regression Analysis With Orange
 
Top profile Call Girls In Satna [ 7014168258 ] Call Me For Genuine Models We ...
Top profile Call Girls In Satna [ 7014168258 ] Call Me For Genuine Models We ...Top profile Call Girls In Satna [ 7014168258 ] Call Me For Genuine Models We ...
Top profile Call Girls In Satna [ 7014168258 ] Call Me For Genuine Models We ...
 
SAC 25 Final National, Regional & Local Angel Group Investing Insights 2024 0...
SAC 25 Final National, Regional & Local Angel Group Investing Insights 2024 0...SAC 25 Final National, Regional & Local Angel Group Investing Insights 2024 0...
SAC 25 Final National, Regional & Local Angel Group Investing Insights 2024 0...
 
Aspirational Block Program Block Syaldey District - Almora
Aspirational Block Program Block Syaldey District - AlmoraAspirational Block Program Block Syaldey District - Almora
Aspirational Block Program Block Syaldey District - Almora
 
Lake Town / Independent Kolkata Call Girls Phone No 8005736733 Elite Escort S...
Lake Town / Independent Kolkata Call Girls Phone No 8005736733 Elite Escort S...Lake Town / Independent Kolkata Call Girls Phone No 8005736733 Elite Escort S...
Lake Town / Independent Kolkata Call Girls Phone No 8005736733 Elite Escort S...
 
Fun all Day Call Girls in Jaipur 9332606886 High Profile Call Girls You Ca...
Fun all Day Call Girls in Jaipur   9332606886  High Profile Call Girls You Ca...Fun all Day Call Girls in Jaipur   9332606886  High Profile Call Girls You Ca...
Fun all Day Call Girls in Jaipur 9332606886 High Profile Call Girls You Ca...
 
Digital Transformation Playbook by Graham Ware
Digital Transformation Playbook by Graham WareDigital Transformation Playbook by Graham Ware
Digital Transformation Playbook by Graham Ware
 
Diamond Harbour \ Russian Call Girls Kolkata | Book 8005736733 Extreme Naught...
Diamond Harbour \ Russian Call Girls Kolkata | Book 8005736733 Extreme Naught...Diamond Harbour \ Russian Call Girls Kolkata | Book 8005736733 Extreme Naught...
Diamond Harbour \ Russian Call Girls Kolkata | Book 8005736733 Extreme Naught...
 
Case Study 4 Where the cry of rebellion happen?
Case Study 4 Where the cry of rebellion happen?Case Study 4 Where the cry of rebellion happen?
Case Study 4 Where the cry of rebellion happen?
 
Top profile Call Girls In Hapur [ 7014168258 ] Call Me For Genuine Models We ...
Top profile Call Girls In Hapur [ 7014168258 ] Call Me For Genuine Models We ...Top profile Call Girls In Hapur [ 7014168258 ] Call Me For Genuine Models We ...
Top profile Call Girls In Hapur [ 7014168258 ] Call Me For Genuine Models We ...
 
Charbagh + Female Escorts Service in Lucknow | Starting ₹,5K To @25k with A/C...
Charbagh + Female Escorts Service in Lucknow | Starting ₹,5K To @25k with A/C...Charbagh + Female Escorts Service in Lucknow | Starting ₹,5K To @25k with A/C...
Charbagh + Female Escorts Service in Lucknow | Starting ₹,5K To @25k with A/C...
 
Introduction to Statistics Presentation.pptx
Introduction to Statistics Presentation.pptxIntroduction to Statistics Presentation.pptx
Introduction to Statistics Presentation.pptx
 
Jual obat aborsi Bandung ( 085657271886 ) Cytote pil telat bulan penggugur ka...
Jual obat aborsi Bandung ( 085657271886 ) Cytote pil telat bulan penggugur ka...Jual obat aborsi Bandung ( 085657271886 ) Cytote pil telat bulan penggugur ka...
Jual obat aborsi Bandung ( 085657271886 ) Cytote pil telat bulan penggugur ka...
 
20240412-SmartCityIndex-2024-Full-Report.pdf
20240412-SmartCityIndex-2024-Full-Report.pdf20240412-SmartCityIndex-2024-Full-Report.pdf
20240412-SmartCityIndex-2024-Full-Report.pdf
 
Top profile Call Girls In Rohtak [ 7014168258 ] Call Me For Genuine Models We...
Top profile Call Girls In Rohtak [ 7014168258 ] Call Me For Genuine Models We...Top profile Call Girls In Rohtak [ 7014168258 ] Call Me For Genuine Models We...
Top profile Call Girls In Rohtak [ 7014168258 ] Call Me For Genuine Models We...
 
Top profile Call Girls In Bihar Sharif [ 7014168258 ] Call Me For Genuine Mod...
Top profile Call Girls In Bihar Sharif [ 7014168258 ] Call Me For Genuine Mod...Top profile Call Girls In Bihar Sharif [ 7014168258 ] Call Me For Genuine Mod...
Top profile Call Girls In Bihar Sharif [ 7014168258 ] Call Me For Genuine Mod...
 
RESEARCH-FINAL-DEFENSE-PPT-TEMPLATE.pptx
RESEARCH-FINAL-DEFENSE-PPT-TEMPLATE.pptxRESEARCH-FINAL-DEFENSE-PPT-TEMPLATE.pptx
RESEARCH-FINAL-DEFENSE-PPT-TEMPLATE.pptx
 
Top profile Call Girls In dimapur [ 7014168258 ] Call Me For Genuine Models W...
Top profile Call Girls In dimapur [ 7014168258 ] Call Me For Genuine Models W...Top profile Call Girls In dimapur [ 7014168258 ] Call Me For Genuine Models W...
Top profile Call Girls In dimapur [ 7014168258 ] Call Me For Genuine Models W...
 

Why advanced population initialization techniques perform poorly in high dimension

  • 1. Why Advanced Population Initialization Techniques Perform Poorly in High Dimension? Borhan Kazimipour Xiaodong Li A.K. Qin
  • 2. Outlines 1. Introduction 2. Background 3. Questions 4. Experiments 5. Results 6. Conclusion SEAL 2014, Dunedin, NZ 2Why Advanced PITs Perform Poorly in HD?
  • 3. Outlines 1. Introduction 2. Background 3. Questions 4. Experiments 5. Results 6. Conclusion SEAL 2014, Dunedin, NZ 3Why Advanced PITs Perform Poorly in HD?
  • 4. Definition of Population Initialization • Definition: – Initialization is the task of generating a set of initial points as potential solutions of an optimization problem. These values are seen as the first position (or distribution) of the individuals in the first generation. • Common Parameters: – Population size – Number of variables or dimensionality (given) – Variables range (given) • Note: In this study our main focus is on continuous techniques capable of generating real-value numbers in continuous spaces. SEAL 2014, Dunedin, NZ 4Why Advanced PITs Perform Poorly in HD?
  • 5. Importance of Population Initialization • Why studying population initialization is important? – Popularity: All population-based algorithms, including EA, need a population initialization module. – “initialize population randomly” is the most widely used expression in EA community! – Variety: Lots of different population initialization techniques are proposed, so far. – About 100 population initialization techniques are proposed so far*. – Effectiveness: Clearly, starting from a good position makes it easier and faster to achieve the aim, than starting from a bad one. – “Advanced initialization techniques can increase the probability of finding global optima, reduce the variation of the final results, decrease the computational costs and improve the solution(s) quality.” * – Inconsistency(!): Some controversy findings have been reported. – “For example, one claimed that the desirable effect of uniformity of initial population is more significant in high dimensions (up to 50 dimensions) while another study, in contrast, claimed that uniform initialization techniques loose their effectiveness in problems of 12 or more dimensions.” * * B. Kazimipour, X. Li, and A. K. Qin. "A review of population initialization techniques for evolutionary algorithms." In Evolutionary Computation (CEC), 2014 IEEE Congress on, pp. 2585-2592. IEEE, 2014. SEAL 2014, Dunedin, NZ 5Why Advanced PITs Perform Poorly in HD?
  • 6. Outlines 1. Introduction 2. Background 3. Questions 4. Experiments 5. Results 6. Conclusion SEAL 2014, Dunedin, NZ 6Why Advanced PITs Perform Poorly in HD?
  • 7. Definitions of Randomness • True Random: – A true random sequence is usually described as a sequence having strong properties such as complete unpredictability, incompressibility and irregularity. – Some believe true random sequences do not exist (theoretical drawback). – There is no tool to proof a given sequence is truly random (empirical drawback). • Computational Random: – A sequence is computationally random if it passes some tests on the properties of true randomness e.g., unpredictability, and incompressibility. • Statistical Random: – A sequence is statistically random if it passes some tests on the statistical (distributional) properties of true random sequences e.g., uniformity. SEAL 2014, Dunedin, NZ 7Why Advanced PITs Perform Poorly in HD? Continuum of RandomnessCompletely Deterministic Truly Random
  • 8. • In this work, we follow the technique proposed in [1] to categorise PITs based on randomness: Does output depend on initial seed? Stochastic Deterministic YES NO Measuring Randomness SEAL 2014, Dunedin, NZ 8Why Advanced PITs Perform Poorly in HD?
  • 9. Categorization based on Randomness SEAL 2014, Dunedin, NZ 9Why Advanced PITs Perform Poorly in HD? Population Initialization Techniques Stochastic Pseudo-Random Number Generator Chaotic Number Generator Deterministic Quasi-Random Sequence Uniform Experimental Design
  • 10. Stochastic vs. Deterministic Stochastic • Definition: – Their results depend on initial seeds. • Properties: – Unpredictable (computationally) – Irregularity • Examples: – Pseudo-Random Number Generator (PRNG) – e.g. WELL, KISS, and Mersenne Twister – Chaotic Number Generator (CNG) – e.g. Tent, Logistic and Sine Deterministic • Definition: – They always generate the same population regardless of any initial seed. • Properties: – Population uniformity is more important than randomness or unpredictability. • Examples: – Quasi-random Sequence – e.g. Sobol, Halton – Uniform Experimental Design – e.g. Latin hypercube, good lattice points and orthogonal design SEAL 2014, Dunedin, NZ Why Advanced PITs Perform Poorly in HD? 10
  • 11. Outlines 1. Introduction 2. Background 3. Questions 4. Experiments 5. Results 6. Conclusion SEAL 2014, Dunedin, NZ 11Why Advanced PITs Perform Poorly in HD?
  • 12. Questions Goal • Research Question: – Why EAs do not receive great benefit from advanced population initialization techniques when dimensionality of problems are very high? • Hypothesis: – The uniformity of population for both simple and advanced techniques drop to the same level when dimensionality grows. SEAL 2014, Dunedin, NZ 12Why Advanced PITs Perform Poorly in HD?
  • 13. Questions… Two experiments Part A (baseline technique) • Goal: Study the trend of population uniformity when generated by popular but simple techniques*. • Research Questions: 1. How much the uniformity of a population can be affected by dimensionality? 2. Is it possible to enhance the uniformity of initial population in high dimensional spaces by increasing the population size? Part B (advanced techniques) • Goal: Compare the performance of advanced initialization techniques with a commonly used technique*. • Research Questions: 1. Can adopting advanced initialization techniques significantly improve population uniformity? 2. How population size affects performance of advanced initializers? SEAL 2014, Dunedin, NZ Why Advanced PITs Perform Poorly in HD? 13 *Random number generators (RNG) are the most widely used initializers in the field of EA.
  • 14. Questions… Quality measures In both parts, we use discrepancy values to measure quality of populations. • Definition of discrepancy: – Literally, discrepancy means non-uniformity. – Technically, discrepancy measures are tools for determining non-uniformity level of a given point set. – Point sets with low discrepancy are those with high level of uniformity. • Variations of discrepancy: – Star L2-discrepancy – Centred L2-discrepancy* – Modified L2-discrepancy – Symmetric L2-discrepancy – Wrap-around L2-discrepancy * Centred L2-discrepancy (CD) is used in this study. SEAL 2014, Dunedin, NZ 14Why Advanced PITs Perform Poorly in HD?
  • 15. Questions… Analytic formulas • L2-discrepancy (D: dimensionality, N: population size, P: population, xi,j: ith value of jth individual) • Centred L2-discrepancy (D: dimensionality, N: population size, P: population, xi,j: ith value of jth individual) SEAL 2014, Dunedin, NZ 15Why Advanced PITs Perform Poorly in HD?
  • 16. Questions… Why we chose discrepancy? Discrepancy measures with analytic formulas are used in this study because: Discrepancy values are not affected by the features of benchmarked problems, employed EAs or their parameters. – Unlike final fitness value and success rate. Discrepancy measures can be easily applied to all kinds of real-value populations. – Unlike DieHard and TestU01 which can only be applied on stochastic population. Discrepancy measures having analytic formulas are faster than similar iterative/recursive algorithms (ideal for large and high dimensional populations). – Unlike early variants of Lp-discrepancy. SEAL 2014, Dunedin, NZ 16Why Advanced PITs Perform Poorly in HD?
  • 17. Outlines 1. Introduction 2. Background 3. Questions 4. Experiments 5. Results 6. Conclusion SEAL 2014, Dunedin, NZ 17Why Advanced PITs Perform Poorly in HD?
  • 18. Experiments Setup • Six population initialization techniques are selected to study. • Three stochastic and three deterministic techniques are included in the experiments. • RNG, which is the most common and simple initializer is chosen as the control method. SEAL 2014, Dunedin, NZ 18Why Advanced PITs Perform Poorly in HD?
  • 19. Experiments… Setup • In both parts: – 20 different dimension sizes are examined (2 ≤ D ≤ 1,000). – 20 different population sizes are examined (10 ≤ N ≤ 10,000). – Each experiment is run for 25 times: – 25 unique initial seeds are used for stochastic techniques – 25 unique sequences are used for deterministic techniques (skip schema) • Part A (baseline technique) – Only performance of RNG is examined in different situations. • Part B (advanced technique) – Performance of advanced techniques are compared with the baseline (RNG). SEAL 2014, Dunedin, NZ 19Why Advanced PITs Perform Poorly in HD?
  • 20. Outlines 1. Introduction 2. Background 3. Questions 4. Experiments 5. Results – Part A – Part B 6. Conclusion SEAL 2014, Dunedin, NZ 20Why Advanced PITs Perform Poorly in HD?
  • 21. Results Part A – Dimensionality effect SEAL 2014, Dunedin, NZ 21Why Advanced PITs Perform Poorly in HD?
  • 22. Results Part A – Dimensionality effect SEAL 2014, Dunedin, NZ 22Why Advanced PITs Perform Poorly in HD? • Discrepancy grows (i.e., uniformity drops) exponentially when the dimensionality increases. – Discrepancy of 10,000 points in 50 dimensions is comparable with the discrepancy of 10 points in 30 dimensions! – 66% growth in dimensionality demands 100,000% increase in population size to recover the uniformity! • For D ≤ 50, a large population size may lessen the undesirable effect of dimensionality (zoomed in the graph)
  • 23. Results Part A – Low dimensions SEAL 2014, Dunedin, NZ 23Why Advanced PITs Perform Poorly in HD?
  • 24. Results Part A – Low dimensions SEAL 2014, Dunedin, NZ 24Why Advanced PITs Perform Poorly in HD? • Population size has no considerable effect on the uniformity of very small-sized problems (D ≤ 10). • For 30 ≤ D ≤ 50, population size has a significant effect on uniformity such that it can be improved 10 to 20 times in the CD scale. • The magnitude of improvements falls rapidly such that increasing population size beyond 1,000 points shows only a minimal improvement.
  • 25. Results Part A – Medium dimensions SEAL 2014, Dunedin, NZ 25Why Advanced PITs Perform Poorly in HD?
  • 26. Results Part A – Medium dimensions SEAL 2014, Dunedin, NZ 26Why Advanced PITs Perform Poorly in HD? • Increase in population size significantly lessens the effect of dimensionality (specially N ≤ 200) • The magnitudes of improvements fall as population grows.
  • 27. Results Part A – High dimensions SEAL 2014, Dunedin, NZ 27Why Advanced PITs Perform Poorly in HD?
  • 28. Results Part A – High dimensions SEAL 2014, Dunedin, NZ 28Why Advanced PITs Perform Poorly in HD? • Uniformity of populations in spaces of above 100 dimensions is so weak that increasing population size from 1,000 to 10,000 cannot recover it. • The feasible and reasonable population size for very large-scale problems (100 ≤ D) is surprisingly less than 1,000 points. • It does not imply N has no effect in D > 100. Instead, it means N must be astronomically large to achieve a significant enhancement. Since evaluating high dimensional populations in that magnitude is currently computationally infeasible, keeping it around 1,000 points is more practical and reasonable.
  • 29. Outlines 1. Introduction 2. Background 3. Questions 4. Experiments 5. Results – Part A – Part B 6. Conclusion SEAL 2014, Dunedin, NZ 29Why Advanced PITs Perform Poorly in HD?
  • 30. Results Part B – Improvement Improvement over common technique: • To compare advanced initialization techniques with a common RNG, we propose a simple formula reflecting relative improvement achieved from each advanced technique: where Pc is the population generated by the control technique (RNG), and Pi is the population produced by the ith advanced initialization technique and CD is centred L2- discrepancy. SEAL 2014, Dunedin, NZ 30Why Advanced PITs Perform Poorly in HD?
  • 31. Results Part B – Low dimensions SEAL 2014, Dunedin, NZ 31Why Advanced PITs Perform Poorly in HD?
  • 32. Results Part B – Low dimensions SEAL 2014, Dunedin, NZ 32Why Advanced PITs Perform Poorly in HD? • Some techniques (TNT and SBL) are successful in improving the common initializer (RNG), although the biggest improvement in 2 ≤ D ≤ 50 is less than 20%. • Some techniques (GLP) are very sensitive to population size, others (SBL) are more stable. • For D ≤ 50 ,with no exception, all techniques work relatively better when population size increases. • Mixed good and bad results can be expected from both categories of initialization techniques*. *B. Kazimipour, X. Li, and A. K. Qin. "Initialization methods for large scale global optimization." In Evolutionary Computation (CEC), 2013 IEEE Congress on, pp. 2750-2757. IEEE, 2013.
  • 33. Results Part B – Medium and High dimensions SEAL 2014, Dunedin, NZ 33Why Advanced PITs Perform Poorly in HD?
  • 34. Results Part B – Medium and High dimensions SEAL 2014, Dunedin, NZ 34Why Advanced PITs Perform Poorly in HD? • All trends converge to one of the three values: 0%, -25% and -80%. • This clearly shows that employing advanced initialization techniques provides no significant improvement in high dimensions, at least in terms of uniformity.
  • 35. Results Part B – Medium and High dimensions SEAL 2014, Dunedin, NZ 35Why Advanced PITs Perform Poorly in HD? • Even increasing population size from 10 to 10,000 does not result in any relative improvement • SBL with 10 and TNT with all population sizes perform almost the same as RNG. • The others, however, perform poorly in comparison with a RNG having the same population size*. * B. Kazimipour, X. Li, and A. K. Qin. "Effects of population initialization on differential evolution for large scale optimization." In Evolutionary Computation (CEC), 2014 IEEE Congress on, pp. 2404-2411. IEEE, 2014.
  • 36. Outlines 1. Introduction 2. Background 3. Questions 4. Experiments 5. Results 6. Conclusion SEAL 2014, Dunedin, NZ 36Why Advanced PITs Perform Poorly in HD?
  • 37. Conclusion What we did • We investigate the reasons that causes advanced population initialization techniques to perform as poor as simple RNG in high dimensional spaces. • We also studied the effect of population size on the quality (uniformity) of the resulting populations. • We studied: – 6 techniques (3 deterministic and 3 stochastic), – 20 dimension sizes (up to 1,000), – 20 population sizes (up to 10,000), – thorough 25 runs. SEAL 2014, Dunedin, NZ 37Why Advanced PITs Perform Poorly in HD?
  • 38. Conclusion What we observed • The uniformity of initial population drops exponentially when dimensionality rises linearly. • Increasing population size up to a computationally feasible bound cannot maintain uniformity (except for some small and medium-sized spaces). • The advanced initializers are as vulnerable to the curse of dimensionality as simple RNG. • Adopting advanced initializers in medium and large-scale spaces does not result in any significant improvement. • Some advanced techniques are even more sensitive to the adverse effect of dimensionality than the simple RNG. SEAL 2014, Dunedin, NZ 38Why Advanced PITs Perform Poorly in HD? We only recommend the use of advanced techniques when the population and dimension sizes are small. In higher dimensional spaces or when the population size is relatively large, no significant improvement is excepted from advanced techniques.
  • 39. Thank you ☺☺☺☺ Any question or comment? 39SEAL 2014, Dunedin, NZ Why Advanced PITs Perform Poorly in HD?