Timeline for How to save and test CNN model on test set after training
Current License: CC BY-SA 4.0
13 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Sep 8, 2018 at 20:07 | comment | added | Green Falcon | I don't know! Are you asking me? :) If your sentence is declarative means you are not overfitting your training data. You should train it more. | |
| Sep 8, 2018 at 14:09 | comment | added | Hunar | I get 0.5 train accuracy? | |
| Sep 8, 2018 at 13:07 | comment | added | Green Falcon | Take a look at here. | |
| Sep 8, 2018 at 13:06 | comment | added | Green Falcon | I guess you have not understood the concept of placeholders yet, try to search for that. Yes, you can and you have to, in order to investigate whether you've overfit the data or not. | |
| Sep 8, 2018 at 13:03 | comment | added | Hunar | train data is used only for training not to check for accuracy. can I do it the same way as validation and test accuracy? or adding it inside train_data loop? | |
| Sep 8, 2018 at 12:59 | comment | added | Green Falcon | What is your train accuracy? | |
| Sep 8, 2018 at 12:58 | comment | added | Hunar | I know it's very small, but why I get 95% of accuracy? have you any idea? | |
| Sep 8, 2018 at 12:56 | comment | added | Green Falcon | Yes, your set is small for training deep models. | |
| Sep 8, 2018 at 12:52 | comment | added | Hunar | ok but I have a problem with that, with only 600 sample of data (360 for training, 120 for each of validation and testing) I get 95% of accuracy for both validation and test, I think this is not real accuracy with this amount of data? | |
| Sep 8, 2018 at 12:49 | vote | accept | Hunar | ||
| Sep 8, 2018 at 12:41 | comment | added | Green Falcon | Yes, exactly. Consider this kind of coding of NNs as a kind of machine that you have tuned its bolts and nuts and you just input something, placeholders, and you get your results. | |
| Sep 8, 2018 at 12:36 | comment | added | Hunar | see the code, you mean something like this? | |
| Sep 8, 2018 at 12:14 | history | answered | Green Falcon | CC BY-SA 4.0 |