5803025047/kelompok3-deeplearning
2
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Add CNN evaluation and visualization (anggota 2)

Browse Source
This commit is contained in:
Exsors
2026-09-26 21:15:12 +07:00
parent 00a821e06d
commit c3f6c5dd8e
9 changed files with 337 additions and 0 deletions
Download Patch File Download Diff File Expand all files Collapse all files
results/reports/classification_report_cnn.csv
+14
View File
@@ -0,0 +1,14 @@
,precision,recall,f1-score,support
0,0.9938775510204082,0.9938775510204082,0.9938775510204082,980.0
1,0.992988606485539,0.9982378854625551,0.9956063268892794,1135.0
2,0.9932038834951457,0.9912790697674418,0.9922405431619786,1032.0
3,0.9940298507462687,0.9891089108910891,0.9915632754342432,1010.0
4,0.9908814589665653,0.9959266802443992,0.9933976637887253,982.0
5,0.9877232142857143,0.992152466367713,0.9899328859060402,892.0
6,0.9937304075235109,0.9926931106471816,0.9932114882506528,958.0
7,0.9893307468477207,0.9922178988326849,0.9907722195240408,1028.0
8,0.9836734693877551,0.9897330595482546,0.9866939611054247,974.0
9,0.9919436052366566,0.9762140733399405,0.984015984015984,1009.0
accuracy,0.9912,0.9912,0.9912,0.9912
macro avg,0.9911382793995284,0.9911440706121668,0.9911311899096779,10000.0
weighted avg,0.9912088120264272,0.9912,0.9911943463223446,10000.0
1 precision recall f1-score support
2 0 0.9938775510204082 0.9938775510204082 0.9938775510204082 980.0
3 1 0.992988606485539 0.9982378854625551 0.9956063268892794 1135.0
4 2 0.9932038834951457 0.9912790697674418 0.9922405431619786 1032.0
5 3 0.9940298507462687 0.9891089108910891 0.9915632754342432 1010.0
6 4 0.9908814589665653 0.9959266802443992 0.9933976637887253 982.0
7 5 0.9877232142857143 0.992152466367713 0.9899328859060402 892.0
8 6 0.9937304075235109 0.9926931106471816 0.9932114882506528 958.0
9 7 0.9893307468477207 0.9922178988326849 0.9907722195240408 1028.0
10 8 0.9836734693877551 0.9897330595482546 0.9866939611054247 974.0
11 9 0.9919436052366566 0.9762140733399405 0.984015984015984 1009.0
12 accuracy 0.9912 0.9912 0.9912 0.9912
13 macro avg 0.9911382793995284 0.9911440706121668 0.9911311899096779 10000.0
14 weighted avg 0.9912088120264272 0.9912 0.9911943463223446 10000.0
results/reports/classification_report_cnn.txt
+19
View File
@@ -0,0 +1,19 @@
CLASSIFICATION REPORT - CNN MNIST (Anggota 2: CNN Evaluation)
Test Accuracy: 99.12% | Test Loss: 0.0268
=================================================================
precision recall f1-score support
0 0.9939 0.9939 0.9939 980
1 0.9930 0.9982 0.9956 1135
2 0.9932 0.9913 0.9922 1032
3 0.9940 0.9891 0.9916 1010
4 0.9909 0.9959 0.9934 982
5 0.9877 0.9922 0.9899 892
6 0.9937 0.9927 0.9932 958
7 0.9893 0.9922 0.9908 1028
8 0.9837 0.9897 0.9867 974
9 0.9919 0.9762 0.9840 1009
accuracy 0.9912 10000
macro avg 0.9911 0.9911 0.9911 10000
weighted avg 0.9912 0.9912 0.9912 10000
results/reports/evaluation_summary_cnn.csv
+2
View File
@@ -0,0 +1,2 @@
test_accuracy_pct,test_loss,total_errors,total_samples,avg_confidence_correct_pct,avg_confidence_wrong_pct,hardest_digit_1,hardest_digit_1_acc_pct,hardest_digit_2,hardest_digit_2_acc_pct
99.12,0.0268,88,10000,99.5341,76.6475,9,97.6214,3,98.9109
1 test_accuracy_pct test_loss total_errors total_samples avg_confidence_correct_pct avg_confidence_wrong_pct hardest_digit_1 hardest_digit_1_acc_pct hardest_digit_2 hardest_digit_2_acc_pct
2 99.12 0.0268 88 10000 99.5341 76.6475 9 97.6214 3 98.9109