%PDF-1.6
%
1 0 obj
<>
endobj
2 0 obj
<>stream
2021-06-18T13:39:08+09:00
2021-06-18T13:39:08+09:00
2021-06-18T13:39:08+09:00
Adobe Acrobat 21.0
application/pdf
Ritsumeikan Univ.KO-374
uuid:7901865e-f2a6-4366-b812-159e1d1fca14
uuid:b144e6b6-bd68-4482-9f8e-da18c87fcc19
Acrobat Web Capture 15.0
endstream
endobj
5 0 obj
<>
endobj
6 0 obj
<>
endobj
3 0 obj
<>
endobj
7 0 obj
<>
endobj
10 0 obj
<>/ProcSet[/PDF/Text]>>/Type/Page>>
endobj
11 0 obj
<>stream
1 g
10 36 574.5 780 re
f
BT
0 g
/TT0 1 Tf
14.25 0 0 14.25 37.7511 788.25 Tm
(A Study of Genetic Algorithms for Searching and Maintaining Building Blo\
cks Based)Tj
6.069 -1.211 Td
(on Symbiotic Evolution Between Individuals and Viruses)Tj
/TT1 1 Tf
12 0 0 12 471.0215 742.5 Tm
(Yoshiaki Sakakura)Tj
-35.543 -2.313 Td
(In this paper, we propose a symbiotic evolutionary genetic algorithm \(S\
EGA\). SEGA aims to enhance)Tj
-1 -1.5 Td
(search efficiency of GA by searching and maintaining building blocks bas\
ed on symbiotic evolution between)Tj
0 -1.5 TD
(individuals and viruses. Various works related to genetic algorithms \(G\
A\) have been studied as one of the)Tj
T*
(stochastic search techniques since GA was proposed. We need to prepare s\
ome individuals which are)Tj
T*
(represented as candidate solutions when we use GA. GA searches solutions\
by Selection, Crossover, and)Tj
T*
(Mutation for the prepared individuals. Recently, many researchers have s\
tudied works which aim to enhance)Tj
T*
(the search efficiency by controlling destructions of building blocks in \
individuals and identifying building)Tj
T*
(blocks in individuals. Our work is classified into these works. Various \
methods have been proposed from)Tj
T*
(these works. SEGA is classified into coevolutionary methods of these met\
hods.)Tj
1 -2.5 Td
(This paper has six chapters. Chapter 1 describes backgrounds and purpose\
s of our works and Chapter 6)Tj
-1 -1.5 Td
(describes conclusions of our works. Chapter 2 describes a procedure, the\
search mechanisms of GA, and)Tj
T*
(definition of a deception problem that GA has a difficulty in searching.\
We also talk about the reason why)Tj
T*
(GA has a difficulty in searching it in Chapter 2. Chapter 3 describes SE\
GA and related works of SEGA. We)Tj
T*
(should deal cautiously with the virus selection and virus evaluation of \
SEGA. Therefore, Chapter 4 describes)Tj
T*
(two kinds of selections and two kinds of evaluations techniques of virus\
es. We compare them each other and)Tj
T*
(discuss about behaviors of SEGAs using them through the function optimiz\
ation experiments . We also show)Tj
T*
(that the most efficient virus selection and evaluation technique in SEGA\
are the sigmoid selection and the)Tj
T*
(independent evaluation. Finally, we apply SEGA using the sigmoid selecti\
on and the independent evaluation)Tj
T*
(to the minimum search problem using some test functions which have disti\
nguishing shape in Chapter 5. We)Tj
T*
(also discuss about the minimum search behaviors, the minimum search effi\
ciency, and building blocks search)Tj
T*
(and maintenance by viruses through comparison SEGA with simple GA and co\
evolutionary GA, which is a)Tj
T*
(related method of SEGA.)Tj
ET
endstream
endobj
12 0 obj
<>
endobj
13 0 obj
(]SS\rrQB.M)
endobj
14 0 obj
<>
endobj
15 0 obj
<>
endobj
18 0 obj
<>
endobj
19 0 obj
<>stream
H\j0E
-EH`iBӖ81Ld8Ɛ]IOf݅a6Gq};o4bQ~ϧߝ۩(Svyc4^^x3w~<|C8ߛ)i>̦21hMv}z?̷T=mtc/S؆/*]+ӼkUU,;?m,*?O'_/-qMKl,dIlccYkQk9bx+Y[JN899;;99;;;;;=LA0S)d
|A0_/\s%[g0_9
(^+rO{tV8+JgYtV8+JgYtV8뚼ahtRWwSk{<uq2
`(>
endstream
endobj
20 0 obj
<>stream
H|yTTGƿۯfdu7<˸&dFGg\ĨJXTE%n"n!ڮ"-1ͤV3$uέ{oUWHڭQ!_~q)+@Q@Ha܈4!^svLLퟒҳV7NgщN@ ]bD sx
?b@Jڨ/Dc!q^gрS"%vTjVPbqplJK@}!uiɧKHttgB>i*ukNt翺u_˯4(9eԡÆ}3bcnOȜäSsN˛>csΛ/Yl
UHk߰q-[m߱s={?pK=vSϜ-ù.^|[m7oB)6)C$Ƒ[cԬ¥H)]ʒrYOUn$.'Y4H~٭Koջ
IAÆwXzآŠS:+=Jo+eE)Q.(VBqC~3SSiiYcf9biVu̱pMx@IQ} 5XV#jfd5G.Vת]D=U#"#DL9AzІm,&$KeŽ&r#hhqv9_Tx^JsnTXb1R+-ʤW*SSra}~S
CXFXQsbT:(1o9U2mJrY