I want to find the minimum of the function in GA to find the values of x. I created a matrix A (70*5) and wrote the function, but after running the GA, the x values are not correct. Is it (fitness function) correct?
function f = simple_fitness(x)
A=[316.7626539 303.3578237 961.6575241 337.2092928 94.6;
137.0433216 190.377636 405.7016489 269.4694834 111.7;
403.6327687 376.3838651 850.0852505 367.4649825 265.3;
266.0557972 168.6624987 514.2429264 128.8305979 237.0;
246.9758995 300.1500678 508.61224 224.2618123 381.6;
338.9456831 328.8217116 961.1093347 373.7654025 190.7;
285.5705355 387.824929 845.8574201 416.2215524 509.7;
285.7322054 382.7136082 376.0398348 232.0847094 163.2;
60.94520673 504.4435148 409.3789356 477.1316661 532.9;
142.450641 180.0703572 380.8780844 180.9126415 232.9;
311.183028 178.9851066 351.3553968 201.6853578 75.4;
231.515141 173.1289053 534.0486375 166.7974282 176.9;
530.8081226 860.8543631 1140.293834 808.0395246 117.3;
191.864544 238.1365323 508.5232394 114.8880213 306.7;
374.1264028 367.0760616 1139.23151 264.0211494 81.4;
302.5424195 349.8440618 416.592934 424.9589701 414.5;
110.7396575 147.3785943 382.4938256 224.153671 80.7;
276.1470996 199.1815368 340.6346724 216.8766959 99.0;
115.1376747 169.7602401 377.4899423 226.7852687 83.8;
146.0560014 182.2560591 364.5698813 169.0558317 118.7;
364.1644753 301.8926488 686.988192 368.1643614 124.0;
382.5270934 248.2526523 740.0773075 408.2022868 222.6;
398.4594989 311.5058112 1036.947919 386.7835726 271.7;
734.6799946 796.466613 1144.232964 738.0273979 270.0;
605.0667735 669.6021081 1160.148843 602.8169181 174.5;
124.5606764 152.3728389 467.923877 158.9729549 143.0;
681.9464481 611.6044791 1445.47763 817.3457145 612.1;
305.9601778 208.4423203 1255.227364 136.5695954 432.7;
1516.07238 1788.088851 1328.598439 1472.918865 1497.5;
2754.602924 2798.928652 1459.85 2614.438019 2800.0;
2269.181412 2345.991037 1215.402366 2202.565876 2321.5;
327.7053159 485.039461 594.7632718 515.2253408 400.0;
372.7843048 476.699024 1174.110871 408.6889789 661.0;
1442.531262 1352.322773 1196.355255 1494.432629 1318.0;
467.3125309 390.7198165 678.3332211 410.6495877 553.0;
343.6173818 540.6854695 1182.815163 394.225525 563.0;
325.958484 214.824587 348.2774766 226.550073 241.0;
262.7118317 191.6752337 377.1790592 170.9270145 257.0;
288.832652 699.036677 632.8128629 847.5446491 885.5;
279.8156452 75.60398239 496.3315453 13.05584332 40.7;
439.2988658 336.2532111 571.5381751 521.0966084 325.0;
279.7976239 36.35408812 514.1520951 63.07298754 80.0;
2030.145387 1963.398022 1213.567174 2090.847461 2025.0;
778.8707556 743.1882188 1161.244574 823.2251663 1147.0;
658.0613959 694.5398724 1201.87063 726.534841 840.0;
884.7159559 873.4369642 1325.528773 630.8994755 853.0;
372.354821 468.3429359 492.2465708 458.0452084 518.0;
266.1491852 211.7637567 362.026381 223.9987793 346.0;
435.2976195 393.1296315 1179.190209 287.5876043 579.0;
745.655746 720.5475937 1379.776763 904.6412664 760.0;
1060.396279 1090.428348 1191.276953 1373.209721 1486.0;
273.2302965 315.530041 789.0791549 172.1676098 86.4;
655.1300643 813.0738677 1189.590962 739.1904211 1296.0;
287.8399372 437.8706863 851.7614621 391.7428388 80.0;
657.5178944 639.4980112 1112.279916 581.6465257 710.0;
662.8299012 707.3526727 1317.64759 576.7251596 689.3;
417.8114895 434.7114516 1019.254879 412.7636494 987.9;
325.0156633 146.9009606 871.3138124 397.2320491 170.6;
119.5162325 106.8205707 480.842798 236.4167669 70.4;
285.4989116 387.1277964 845.8574201 416.089543 727.7;
291.6529578 297.9626571 906.2924125 195.9043984 265.6;
341.9063567 263.0328678 1190.155048 133.7030761 169.6;
2667.871352 2803.986485 1459.85 2851.736185 2800.0;
2768.952289 2875.145392 1459.85 2904.32357 2879.0;
323.2200114 373.5511513 638.8234642 315.8957268 809.0;
349.0296173 249.1168443 416.4372719 77.75868469 270.0;
374.2546852 352.3715148 1236.300428 478.1740085 147.0;
107.9588485 163.3857133 363.348985 230.7611383 138.0;
971.5750297 983.9590704 1177.419044 1202.403727 1371.3;
311.7549784 274.7452781 1098.561085 251.1600602 90.7;]
f=symsum((1/70)*((x(1)*A(n,1)+x(2)*A(n,2)+x(3)*A(n,3)+x(4)*A(n,4)-A(n,5)).^2),n,1,70);
Are you sure you want a series sum here?
I think what you are looking for could be:
function f = simplefun(x)
A = [];
f = 1/70 * sum(A(:,1:4)*x(:) - A(:,5));
end
assuming that x is a 4-vector. Using that gives me
>> x = ga(@simplefun,4)
Optimization terminated: maximum number of generations exceeded.
x =
1.0e+03 *
-1.4870 -1.6133 -1.6435 -1.5389
Hope it helps.
(note that this can be optimized, since, e.g., there is no need to recompute the sum of the last column on every iteration.)
Ok. That's (usually) not up to the fitness function to define - it should be honored by the optimization algorithm. How about this:
x = ga(@simplefun,4,[],[],ones(1,4),1,zeros(4,1),ones(4,1))
Optimization terminated: average change in the fitness value less than options.TolFun.
x =
0.9984 0.0005 0.0000 0.0001
Dear R. Barzegar,
I think you should specify what you want to minimise, because you say that your mathematical function to minimise is:
f=sigma(1/70)*((A1*x1)+(A2*x2)+(A3*x3)+(A4*x4)-A5)2
but you provide 70 values of each Ai. If you want to minimise the sum of all them your target function should be the proposed by Jacob, but if it is not, please specify it.
On the other hand, Matlab GA considers LB
DEar Ferran Torrent Fontbona,
I want to minimize the sum of all them indeed the lowest value for function.
Then I agree with Jacob that your fitness function should be:
f = 1/70 * sum(A(:,1:4)*x(:) - A(:,5));
Then you should choose how to deal with constraints:
- Badges
- Science topic