Timeline for Detect if my ANN model is overfitted
Current License: CC BY-SA 4.0
8 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Nov 19, 2019 at 9:23 | comment | added | Gaurav Roy | great, thanks a lot. | |
| Nov 19, 2019 at 9:19 | comment | added | Piotr Rarus | Do projects. People learn by experience. Watching tutorials will get you nowhere. When it comes to DNN it's nice to have someone holding your hand and start from basics. This is great book. Cannot recommend it more. | |
| Nov 19, 2019 at 9:17 | comment | added | Gaurav Roy | basic question , on how to learn more of machine learning and CNN? like any tips? | |
| Nov 19, 2019 at 9:10 | comment | added | Piotr Rarus |
Ups, got my math wrong. Constant not-fraud model would give you 99.3% accuracy.
|
|
| Nov 19, 2019 at 9:08 | comment | added | Gaurav Roy | oh okay, great. thanks. got to know something new. will ask if any doubgs | |
| Nov 19, 2019 at 8:57 | comment | added | Piotr Rarus |
Your model overfits to over-represented class. You can just output not-fraud all the time and you'll get 83% accuracy. Accuracy is counted globally and doesn't account for imbalance. Check this kernel and this one. Before you jump into DNN you should first explore more classical models like random forest or xgboost, they're much faster to fit. This data set is rather tough and you won't get away with just simple rescaler.
|
|
| Nov 19, 2019 at 8:45 | comment | added | Gaurav Roy | okay , so why didnt my model learn anything? didnt get that part? i should have a balanced dataset for that? and how to tackle it? | |
| Nov 19, 2019 at 8:34 | history | answered | Piotr Rarus | CC BY-SA 4.0 |