Commit
•
3d47f3a
1
Parent(s):
47e09f6
Update README.md
Browse files
README.md
CHANGED
|
@@ -40,6 +40,63 @@ this model will identify diabetic retinopathy on a scale of 0 to 4:
|
|
| 40 |
## Evaluation
|
| 41 |
Training accuracy - trained for 50 epochs, reaching 83% accuracy within our training data
|
| 42 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 43 |
|
| 44 |
|
| 45 |
We submitted our model for validation to the [APTOS 2019 Blindness Detection Competition](https://www.kaggle.com/competitions/aptos2019-blindness-detection/submissions#),
|
|
|
|
| 40 |
## Evaluation
|
| 41 |
Training accuracy - trained for 50 epochs, reaching 83% accuracy within our training data
|
| 42 |
|
| 43 |
+
| Epoch | Train Loss | Valid Loss | Accuracy | Error Rate | Time |
|
| 44 |
+
|-------|------------|------------|----------|------------|-------|
|
| 45 |
+
| 0 | 1.271288 | 1.351223 | 0.665301 | 0.334699 | 03:47 |
|
| 46 |
+
| 1 | 1.013268 | 0.742499 | 0.741803 | 0.258197 | 04:12 |
|
| 47 |
+
| 2 | 0.806825 | 0.687152 | 0.754098 | 0.245902 | 03:42 |
|
| 48 |
+
| 0 | 0.631816 | 0.533298 | 0.789617 | 0.210383 | 04:22 |
|
| 49 |
+
| 1 | 0.537469 | 0.457713 | 0.829235 | 0.170765 | 04:23 |
|
| 50 |
+
| 2 | 0.498419 | 0.515875 | 0.810109 | 0.189891 | 04:20 |
|
| 51 |
+
| 3 | 0.478353 | 0.511856 | 0.815574 | 0.184426 | 04:13 |
|
| 52 |
+
| 4 | 0.459457 | 0.475843 | 0.801913 | 0.198087 | 04:17 |
|
| 53 |
+
| 5 | 0.416427 | 0.491665 | 0.826503 | 0.173497 | 03:27 |
|
| 54 |
+
| 6 | 0.415657 | 0.491500 | 0.826503 | 0.173497 | 03:24 |
|
| 55 |
+
| 7 | 0.453110 | 0.549434 | 0.796448 | 0.203552 | 03:27 |
|
| 56 |
+
| 8 | 0.445712 | 0.498385 | 0.834699 | 0.165301 | 03:25 |
|
| 57 |
+
| 9 | 0.433412 | 0.538176 | 0.801913 | 0.198087 | 03:20 |
|
| 58 |
+
| 10 | 0.428817 | 0.489114 | 0.834699 | 0.165301 | 03:24 |
|
| 59 |
+
| 11 | 0.439062 | 0.737071 | 0.785519 | 0.214481 | 03:20 |
|
| 60 |
+
| 12 | 0.416182 | 0.532924 | 0.808743 | 0.191257 | 03:20 |
|
| 61 |
+
| 13 | 0.385860 | 0.482569 | 0.811475 | 0.188525 | 03:26 |
|
| 62 |
+
| 14 | 0.405143 | 0.527165 | 0.807377 | 0.192623 | 03:27 |
|
| 63 |
+
| 15 | 0.374315 | 0.539998 | 0.831967 | 0.168033 | 03:22 |
|
| 64 |
+
| 16 | 0.372636 | 0.468733 | 0.841530 | 0.158470 | 03:22 |
|
| 65 |
+
| 17 | 0.339173 | 0.617834 | 0.800546 | 0.199454 | 03:24 |
|
| 66 |
+
| 18 | 0.329398 | 0.557657 | 0.818306 | 0.181694 | 03:24 |
|
| 67 |
+
| 19 | 0.331022 | 0.536041 | 0.808743 | 0.191257 | 03:23 |
|
| 68 |
+
| 20 | 0.290718 | 0.532082 | 0.836066 | 0.163934 | 03:23 |
|
| 69 |
+
| 21 | 0.264355 | 0.541103 | 0.837432 | 0.162568 | 03:22 |
|
| 70 |
+
| 22 | 0.241962 | 0.606261 | 0.806011 | 0.193989 | 03:25 |
|
| 71 |
+
| 23 | 0.225110 | 0.634152 | 0.814208 | 0.185792 | 03:21 |
|
| 72 |
+
| 24 | 0.204949 | 0.686106 | 0.821038 | 0.178962 | 03:21 |
|
| 73 |
+
| 25 | 0.193531 | 0.706426 | 0.804645 | 0.195355 | 03:24 |
|
| 74 |
+
| 26 | 0.169020 | 0.653785 | 0.827869 | 0.172131 | 03:22 |
|
| 75 |
+
| 27 | 0.158583 | 0.727126 | 0.810109 | 0.189891 | 03:21 |
|
| 76 |
+
| 28 | 0.143771 | 0.652805 | 0.834699 | 0.165301 | 03:20 |
|
| 77 |
+
| 29 | 0.140031 | 0.667536 | 0.829235 | 0.170765 | 03:20 |
|
| 78 |
+
| 30 | 0.129302 | 0.665816 | 0.829235 | 0.170765 | 03:21 |
|
| 79 |
+
| 31 | 0.119224 | 0.777733 | 0.810109 | 0.189891 | 03:21 |
|
| 80 |
+
| 32 | 0.102287 | 0.736923 | 0.827869 | 0.172131 | 03:21 |
|
| 81 |
+
| 33 | 0.100752 | 0.709455 | 0.826503 | 0.173497 | 03:21 |
|
| 82 |
+
| 34 | 0.080087 | 0.710993 | 0.845628 | 0.154372 | 03:21 |
|
| 83 |
+
| 35 | 0.064930 | 0.768080 | 0.831967 | 0.168033 | 03:20 |
|
| 84 |
+
| 36 | 0.066414 | 0.787695 | 0.814208 | 0.185792 | 03:19 |
|
| 85 |
+
| 37 | 0.058803 | 0.788034 | 0.830601 | 0.169399 | 03:22 |
|
| 86 |
+
| 38 | 0.063010 | 0.758910 | 0.836066 | 0.163934 | 03:22 |
|
| 87 |
+
| 39 | 0.061156 | 0.792248 | 0.834699 | 0.165301 | 03:24 |
|
| 88 |
+
| 40 | 0.047776 | 0.751294 | 0.825137 | 0.174863 | 03:25 |
|
| 89 |
+
| 41 | 0.052572 | 0.763859 | 0.838798 | 0.161202 | 03:23 |
|
| 90 |
+
| 42 | 0.035978 | 0.797310 | 0.836066 | 0.163934 | 03:29 |
|
| 91 |
+
| 43 | 0.037231 | 0.801716 | 0.826503 | 0.173497 | 03:26 |
|
| 92 |
+
| 44 | 0.031092 | 0.819678 | 0.831967 | 0.168033 | 03:25 |
|
| 93 |
+
| 45 | 0.033036 | 0.810405 | 0.836066 | 0.163934 | 03:26 |
|
| 94 |
+
| 46 | 0.030294 | 0.798600 | 0.840164 | 0.159836 | 03:22 |
|
| 95 |
+
| 47 | 0.027893 | 0.821619 | 0.840164 | 0.159836 | 03:21 |
|
| 96 |
+
| 48 | 0.024947 | 0.800241 | 0.840164 | 0.159836 | 03:21 |
|
| 97 |
+
| 49 | 0.027916 | 0.803851 | 0.838798 | 0.161202 | 03:26 |
|
| 98 |
+
|
| 99 |
+
|
| 100 |
|
| 101 |
|
| 102 |
We submitted our model for validation to the [APTOS 2019 Blindness Detection Competition](https://www.kaggle.com/competitions/aptos2019-blindness-detection/submissions#),
|