Timeline for opencv neural network, incorrect predict
Current License: CC BY-SA 3.0
Post Revisions
16 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Apr 8, 2016 at 15:35 | vote | accept | colla | ||
| Apr 8, 2016 at 15:02 | comment | added | Lilith Daemon | What transfer function are you using and what is your sample size? (And your eta/alpha) You could possibly be over/undertraining it. For a simple test try creating a XOR gate with the neural net, and then working as MSalters on the MNIST dataset. To save some time, look at stackoverflow.com/questions/8286668/how-to-read-mnist-data-in-c for how to read the MNIST in c++. | |
| Apr 8, 2016 at 9:20 | history | edited | colla | CC BY-SA 3.0 |
added 424 characters in body
|
| Apr 8, 2016 at 6:57 | comment | added | MSalters | Using the MNIST set of handwritten digits is a good way to check your own implementation - for that data set it's well-known how good different algorithms should perform. | |
| Apr 8, 2016 at 1:41 | answer | added | Aenimated1 | timeline score: 3 | |
| Apr 7, 2016 at 23:53 | comment | added | sturkmen | maybe this link will be useful. | |
| Apr 7, 2016 at 21:57 | review | Close votes | |||
| Apr 11, 2016 at 0:03 | |||||
| Apr 7, 2016 at 21:28 | answer | added | Carafini | timeline score: 0 | |
| Apr 7, 2016 at 21:27 | history | edited | colla | CC BY-SA 3.0 |
deleted 11 characters in body
|
| S Apr 7, 2016 at 21:24 | history | suggested | Sabuncu | CC BY-SA 3.0 |
Made images inline.
|
| Apr 7, 2016 at 21:21 | review | Suggested edits | |||
| S Apr 7, 2016 at 21:24 | |||||
| Apr 7, 2016 at 21:16 | history | edited | Rao | CC BY-SA 3.0 |
Improved formatting.
|
| Apr 7, 2016 at 21:13 | history | edited | colla | CC BY-SA 3.0 |
added 20 characters in body
|
| Apr 7, 2016 at 21:11 | review | First posts | |||
| Apr 7, 2016 at 21:16 | |||||
| S Apr 7, 2016 at 21:07 | review | Triage | |||
| Apr 7, 2016 at 21:52 | |||||
| S Apr 7, 2016 at 21:07 | history | asked | colla | CC BY-SA 3.0 |