Tengo datos de entrenamiento como los siguientes :
1 1:128.319 2:71.4336 3:130.255 4:292.948 5:96.3541 6:71.942 7:136.189 8:71.5032 9:148.21 10:304.011 11:90.3781 12:99.4496 13:164.268 14:138.892 15:220.871 16:139.198 17:151.709
1 1:118.431 2:97.5874 3:92.0313 4:242.383 5:32.5916 6:91.4476 7:122.88 8:86.6046 9:104.173 10:278.208 11:100.356 12:44.0638 13:159.678 14:163.307 15:181.011 16:83.5148 17:96.1238
1 1:137.925 2:103.664 3:129.569 4:280.765 5:48.4616 6:93.4706 7:104.143 8:87.889 9:125.967 10:284.253 11:80.2913 12:51.7028 13:125.539 14:162.068 15:203.335 16:175.856 17:177.638
1 1:134.952 2:141.661 3:247.987 4:288.032 5:41.1806 6:109.573 7:117.331 8:99.5418 9:159.184 10:232.882 11:78.1744 12:75.0752 13:168.181 14:132.876 15:204.355 16:107.003 17:118.515
0 1:115.603 2:61.5781 3:132.382 4:293.832 5:65.9111 6:106.214 7:103.853 8:116.03 9:142.142 10:293.42 11:49.4609 12:97.0318 13:157.035 14:146.374 15:179.402 16:135.535 17:125.605
0 1:128.367 2:68.9784 3:83.4542 4:233.359 5:71.6529 6:69.5444 7:134.49 8:64.3986 9:87.4766 10:282.327 11:87.1273 12:61.7704 13:136.858 14:131.094 15:164.514 16:68.9703 17:70.6821
0 1:109.747 2:85.5685 3:139.116 4:309.536 5:53.0768 6:93.8608 7:100.835 8:87.8225 9:131.276 10:277.85 11:93.1516 12:49.7809 13:163.615 14:148.949 15:236.783 16:148.671 17:186.764
0 1:112.509 2:79.2363 3:153.702 4:271.478 5:74.3454 6:86.4557 7:86.2066 8:84.2486 9:156.918 10:254.445 11:74.8842 12:75.8982 13:112.379 14:108.054 15:164.201 16:84.1002 17:102.252
0 1:141.522 2:78.2754 3:127.487 4:363.352 5:91.2171 6:81.9828 7:231.043 8:113.83 9:229.079 10:437.796 11:122.712 12:115.435 13:178.213 14:155.286 15:188.521 16:131.901 17:124.573
Y cuando ejecuto el siguiente código:
[true_label, data_inst] = libsvmread('C:\Users\fypstudent\Documents\MATLAB\Training_Data\OVA_E0.txt');
model = svmtrain(true_label, data_inst, '-c 1 -g 0.07 -t 0 -b 1');
[true_label, data_inst] = libsvmread('C:\Users\fypstudent\Documents\MATLAB\Training_Data\OVA_E0.txt');
[predict_label, accuracy, prob_estimates] = svmpredict(true_label, data_inst, model, '-b 1');
Obtengo las siguientes probabilidades para los datos que proporcioné:
0.145614770432359 0.854385229567641
0.146701861245572 0.853298138754428
0.146126482529063 0.853873517470937
0.148072155086074 0.851927844913926
0.147143517011699 0.852856482988301
0.145711409800836 0.854288590199164
0.144862609974806 0.855137390025194
0.145741700456742 0.854258299543258
0.146935964238630 0.853064035761370
Me gustaría saber qué indican estas probabilidades. También obtengo una probabilidad > 0.5 para etiquetas falsas, ¿alguien puede ayudarme a entender esto?