method 2.txt

GA sources code in C++

%#################################################%
帪宯楍僨乕僞偺帪娫揑娭學傪峫椂偟偨曽朄Method of considering time relation of the longitudinal data
%#################################################%

慜愡傑偱偺曽朄偱偼丆帪宯楍僨乕僞偵娭偡傞帪娫揑側娭學偼峫椂偝傟偰偄側偐偭偨丏偮傑傝丆
1擭慜偺僨乕僞偲2擭慜偺僨乕僞偵帪娫揑側堘偄偼側偔丆摨摍偵峫偊傜傟偰偄偨偺偱偁傞丏
In the method to the foregoing paragraph, a time relation concerning the longitudinal 
data was not considered. In a word, a time difference was not, and was thought 
equally by the data before one year and the data two years ago. 

偦偙偱杮愡偱偼丆帪宯楍僨乕僞偺帪娫揑側娭學傪夞婣學悢傪梡偄偰昞傢偟丆儕僗僋嵟彫丆
婜懸抣嵟戝偲偡傞擇栚揑寁夋栤戣偲偟偰掕幃壔偡傞曽朄傪採埬偡傞丏
Then, a time relation of the longitudinal data is shown by using the regression 
coefficient, and it proposes the method of the formulation as two purpose plan 
problem assumed to be a minimum risk and the maximum expected value in this 
chapter. 

丂

%昞丂戞i梫場偺僨乕僞(庤朄丂II)
\begin{table}[htbp]
\caption{戞$i$梫場偺僨乕僞}
\begin{center}
 \begin{tabular}{|c|c||c|} \hline
帪娫~~$t$ & 僨乕僞偺幚愌抣~~$\displaystyle y_i^{(t)}$ & 僨乕僞偺悇掕抣~~$\displaystyle Y_i^{(t)}$ \\ \hline \hline
$1$ & $\displaystyle y_i^{(1)}$ & $\displaystyle Y_i^{(1)}$ \\ \hline
$2$ & $\displaystyle y_i^{(2)}$ & $\displaystyle Y_i^{(2)}$ \\ \hline
$\vdots$ & $\vdots$ & $\vdots$ \\ \hline
$N$ & $\displaystyle y_i^{(N)}$ & $\displaystyle Y_i^{(N)}$ \\ \hline
 \end{tabular}
\end{center}
\label{youin1}
\end{table}

帪娫揑側娭學傪峫椂偡傞偨傔偵丆戞$i$梫場偺僨乕僞偺悇掕抣傪埲壓偺幃偱昞傢偡偙偲偵偡傞丏

\begin{eqnarray}
Y_i^{(t)}=a_i+b_it
\end{eqnarray}

\begin{tabular}{lcl}
$Y_i^{(t)}$ &:& 戞$i$梫場偺$t$帪揰偱偺僨乕僞偺悇掕抣 \\
$t$ &:& 帪娫丂$(1,2,\cdots,N)$ \\
$N$ &:& 僨乕僞偺婜娫悢 \\
$a_i$ &:& 戞$i$梫場偺夞婣掕悢 \\
$b_i$ &:& 戞$i$梫場偺僨乕僞偺帪娫偵懳偡傞夞婣學悢 \\
\end{tabular}

丂

偮傑傝丆怴偟偄僨乕僞傎偳廳梫搙偑崅偄偙偲傪堄枴偟偰偄傞丏

偙偙偱丆$\displaystyle \sum_{t=1}^N \{ y_i^{(t)}-Y_i^{(t)} \} ^2$傪嵟彫偵偡傞傛偆偵丆$a_i,b_i$傪媮傔傞~\cite{reidai_b1}丏

丂

偟偨偑偭偰丆儕僗僋嵟彫偍傛傃婜懸抣嵟戝偲偡傞擇栚揑寁夋栤戣偲偟偰埲壓偺傛偆偵掕幃壔偱偒傞丏

\begin{flushleft}
乵掕幃5乶
\end{flushleft}
\vskip -4mm
\begin{eqnarray}
\hbox to 1cm{min.} & & V(\mbox{\boldmath $x$}) = \frac{1}{Q^2} \sum_{i=1}^n \sum_{j=1}^n \sigma_{ij} b_i x_i b_j x_j \\
\hbox to 1cm{max.} & & E(\mbox{\boldmath $x$}) = \frac{1}{Q} \sum_{i=1}^n \mu_i b_i x_i \\
\hbox to 1cm{s.t.} & & \sum_{i=1}^n x_i = Q \\
\null              & & x_i \ge 0 丂丂(i=1,2,乧,n)
\end{eqnarray}

偨偩偟丆$b_i$偼戞$i$梫場偺僨乕僞偺帪娫偵懳偡傞夞婣學悢丆$\sigma_{ij}$偼戞$i$梫場偲戞$j$梫場偲偺嫟暘嶶丆$\mu_i$偼戞$i$梫場偺婜懸抣丆$Q$偼攝暘壜擻検偱偁傞丏傑偨丆$x_i$偼戞$i$梫場傊偺攝暘検偱偁傝丆師椺偺傛偆偵揔梡偡傞栤戣偵傛偭偰偼$x_i$偼惍悢偲側傞丏

丂

偙偺庤朄傪梡偄偰丆奺抧堟偵壗恖偺恖堳傪攝暘偡傞偐傪寛掕偡傞栤戣傪夝偔偙偲傪帋傒傞丏
It is tried to solve the problem of deciding how many number of men to distribute 
to various places by using this technique. 

栤戣愝掕偲偟偰丆戝嶃晎壓偵偁傞帺摦幵斕攧夛幮偺恖堳攝抲栤戣傪庢傝埖偆丏偙偺帺摦幵斕攧夛幮偼丆
11偺斕攧抧堟偱僙乕儖僗偑峴傢傟偰偍傝丆怴婯偵嵦梡偟偨恖堳傪偄偐偵攝抲偟偰偄偔偐傪栤戣偲偟
偰偄傞丏偙偺偲偒丆攝暘壜擻検Q傪3恖偐傜7恖偲偟偰偄傞丏偙傟偼丆恖審旓偵娭學偟偨悢抣偱
偁傞丏梡偄傞僨乕僞偼丆暯惉2擭1寧偐傜暯惉5擭12寧傑偱偺48婜娫偵傢偨傞奺抧堟偛偲偺恖堳1恖摉偨
傝偺攧忋崅偺帪宯楍(扨埵:10枩墌)偱偁傝丆偦傟傪婎偵偟偰夞婣學悢丆暘嶶嫟暘嶶峴楍丆婜懸抣傪
嶼弌偟堄巚寛掕幰偺婓媮悈弨偵墳偠偨枮懌夝偺摫弌傪峴偭偨丏奺抧堟偵娭偡傞夞婣學悢傪昞偵丆婜
懸抣偍傛傃暘嶶嫟暘嶶偵娭偟偰偼丆昞12偍傛傃昞13偵帵偡.
The staff assignment problem of the car sales companies under Osaka Prefecture is handled as 
a problem setting. To be for sales to be done in the selling area of 11, and how the problem 
of this car sales companies arranged the number of men newly adopted?At this time, three 
seven people assume amount Q that can be distributed. This is a numerical value that relates 
to the labor cost. The data used was time series of sales the number of men person of each 
various places (The unit: 100,000 yen) for 48 periods from January, 1990 to December, 1993, 
and the regression coefficient, the decentralized covariance procession, and the expected 
value were calculated based on it and the satisfaction solution was derived according to 
decision-maker's desire level. The regression coefficient concerning various places is shown 
in Table 12 and Table 13 in the table for an expected value and a decentralized covariance. 

 栚昗儕僗僋偵懳偡傞昁梫儗儀儖傪0.017丆廫暘儗儀儖傪0.007偲偡傞丏栚昗婜懸抣偵懳偡傞昁梫
儗儀儖傪0.55丆廫暘儗儀儖傪0.75偲偡傞丏V_M偍傛傃E_M偼丆V_M = (0.017+0.007)/2 = 0.012丆
E_M = (0.55+0.75)/2 = 0.65偲側傞丏傑偨丆儊儞僶乕僔僢僾娭悢偺宍忬偵娭偡傞僷儔儊乕僞偼丆
alpha_V=920$偍傛傃alpha_E=46偱偁傞丏
The level is assumed to be 0.007 enough ..a necessary level to the target risk.. ..0.017... 
The level is assumed to be 0.75 enough ..a necessary level to the expected value of the 
target.. ..0.55... 2 (0.55+0.75)/ 2 (0.017+0.007)/ V_M and E_M are V_M = = 0.012 and 
E_M = = It becomes 0.65. Moreover, the parameter concerning the shape of the 
membership function is alpha_V=920$ and alpha_E=46. 



偙偙偱丆堚揱揑傾儖僑儕僘儉偺僷儔儊乕僞偲偟偰丆廤抍僒僀僘傪150偲偟丆搼懣棪傪10%丆
岎嵆棪傪60%丆撍慠曄堎棪傪$3$\%丆嵟戝悽戙悢傪200偲偟偨丏偙偺僷儔儊乕僞乕偱丆
僔儈儏儗乕僔儑儞傪10夞峴偭偨寢壥傪昞偵帵偡丏傑偨丆偙偺応崌偺
廂懇忬嫷偺堦椺傪恾偵帵偡丏
Here..genetic algorithm..parameter..group..size..do..selection..rate..
intersection..rate..mutation rate..maximum..generation..number..do.The result of 
the simulation ten times is shown in the table by this parameter. Moreover, one 
example of the settling situation in this case is shown in figure. 


%昞丂夞婣暘愅偺寢壥(庤朄丂II)Result of regression analysis
\begin{table}[htbp]
\caption{夞婣暘愅偺寢壥}
\begin{center}
 \begin{tabular}{|c||c|c|} \hline
抧堟    & $\displaystyle a_i$ & $\displaystyle b_i$ \\ \hline \hline
抧堟丂1 &  9.5478 & 0.02652 \\ \hline
抧堟丂2 & 10.0508 & 0.04210 \\ \hline
抧堟丂3 &  9.9030 & 0.05642 \\ \hline
抧堟丂4 & 10.5721 & 0.00926 \\ \hline
抧堟丂5 & 65.0652 & 0.05606 \\ \hline
抧堟丂6 & 13.3404 & 0.03896 \\ \hline
抧堟丂7 & 10.6266 & 0.04196 \\ \hline
抧堟丂8 & 10.3827 & 0.06692 \\ \hline
抧堟丂9 &  8.7043 & 0.02633 \\ \hline
抧堟丂10 & 8.7118 & 0.06309 \\ \hline
抧堟丂11 & 8.9334 & 0.04004 \\ \hline
 \end{tabular}
\end{center}
\label{kaiki1}
\end{table}

\newpage

%乵昞12,13 傪憓擖偟偰壓偝偄乶
\addtocounter{page}{1}
\addtocounter{table}{2}

%昞丂堚揱揑傾儖僑儕僘儉偱偺僔儈儏儗乕僔儑儞偺寢壥(庤朄丂II)
\begin{table}[htbp]
\caption{堚揱揑傾儖僑儕僘儉偱偺僔儈儏儗乕僔儑儞偺寢壥}
\begin{center}
 \begin{tabular}{|c||r|} \hline
暯嬒扵嶕悢 & $3335.1$ \\ \hline
暯嬒扵嶕斾棪 & $0.016\%$ \\ \hline
暯嬒敪尒悽戙悢 & $10.6$ \\ \hline
嵟揔夝摓払夞悢 & $10$ \\ \hline
 \end{tabular}
\end{center}
\label{ga_kekka2}
\end{table}

%恾丂廂懇忬嫷偺堦椺(庤朄丂II)
\begin{figure}[htbp]
\begin{center}
  \input{tekioug2.tex} \\
\end{center}
\caption{廂懇忬嫷偺堦椺}
\label{syuusoku2}
\end{figure}

寢壥偲偟偰丆抧堟2, 6偍傛傃10偵1恖丆抧堟8偵3恖丆攝抲偟偨応崌偑丆嵟傕婣懏搙偺崅偄夝丆偮傑傝枮懌夝偑摼傜傟偨丏
They were the people of one of ..region.. two as the result in 6 and 10, and they were three people in 
region 8, and the solutions with the highest belonging level in the arranged case ,in a word, obtaining 
the satisfaction solution. 

%昞丂摼傜傟偨婣懏搙丆栚昗儕僗僋偍傛傃栚昗婜懸抣(庤朄丂II)
\begin{table}[htbp]
\caption{摼傜傟偨婣懏搙丆栚昗儕僗僋偍傛傃栚昗婜懸抣}
\begin{center}
 \begin{tabular}{|c||r|} \hline
  婣懏搙$\lambda$  &  $0.8043$ \\ \hline
  栚昗儕僗僋       &  $0.01027$ \\ \hline
  栚昗婜懸抣       &  $0.6807$ \\ \hline
 \end{tabular} \\

丂

拲).攧忋崅/恖堳~(10枩墌)
\end{center}
\label{r2_kekka2}
\end{table}