Commit graph

Select branches

Hide pull requests

digits

main

#1

Mono Color

• 6a23b6a4ae Limited log window size main mandlm 2015-11-16 17:49:40 +0100
• c5fbe764a6 Limited ErrorPlotter buffer size mandlm 2015-11-16 17:27:42 +0100
• 0f7c617f9c Back to learning all samples Michael Mandl 2015-11-15 16:09:25 +0100
• 5e32724b1a Layers average their output sums Michael Mandl 2015-11-15 16:08:49 +0100
• 44760d0402 Debugging-setup with a single training sample to check the lack of learning progress Michael Mandl 2015-11-01 15:01:43 +0100
• 1d343a079a Moved MNIST sample to QImage conversion to MNIST loader Michael Mandl 2015-11-01 15:00:32 +0100
• eecd7a0fe6 Querying of a single MNIST sample Michael Mandl 2015-11-01 14:49:02 +0100
• ab9dcfbd35 Clean learner-thread termination on program exit Michael Mandl 2015-11-01 12:36:23 +0100
• 1a0d2b9ea7 Merge pull request #1 from mandlm/digits mandlm 2015-10-31 15:08:01 +0100
• 8d01edb7a1 Removed naked double * calls digits Michael Mandl 2015-10-31 14:59:10 +0100
• cd1101dfe2 Learning from and displaying of digit samples Michael Mandl 2015-10-31 14:58:49 +0100
• d98ec63fbd Completed loading of MNIST Databases mandlm 2015-10-31 12:28:35 +0100
• 87e267d65f Implemented parsing of MNIST files Michael Mandl 2015-10-29 16:00:58 +0100
• 83b4562a29 Removed useless training images, added MNIST database instead Michael Mandl 2015-10-29 13:06:30 +0100
• 5778afa121 Display training digits in UI while training Michael Mandl 2015-10-27 18:20:50 +0100
• f9be5ca717 Load and train handwritten digits Michael Mandl 2015-10-27 15:33:54 +0100
• 650b4be9fc Added forward-feeding of double-ptrs Michael Mandl 2015-10-27 15:33:10 +0100
• ec654c0e5f Improved icon visibility Michael Mandl 2015-10-27 15:32:46 +0100
• 81d1f54c98 Started loader-class for 8x8 pixel digit training data mandlm 2015-10-26 22:05:50 +0100
• 80da55a2b8 Added handwritten digits as training data mandlm 2015-10-26 21:34:13 +0100
• aa61792911 Fixed a performance problem with a QListWidget while inserting and scrolling mandlm 2015-10-26 20:51:36 +0100
• 37b5153d6a Cleaned up code to work with MinGW/GCC Kit mandlm 2015-10-26 20:34:26 +0100
• 1e716979a9 Reduced the progress update messages to take load from the UI, implemented load-or-create in test funtion Michael Mandl 2015-10-26 19:50:01 +0100
• 9bb927d2d2 Fixed a problem with multiple connects Michael Mandl 2015-10-26 17:54:06 +0100
• 22dba2cb27 Added time measurement Michael Mandl 2015-10-26 09:19:48 +0100
• 24b6969ab7 Added a error-plotter widget to NeuroUI Michael Mandl 2015-10-26 08:41:59 +0100
• b899c6f55e Replaced a few int with size_t to fix 64 bit warnings. Michael Mandl 2015-10-26 07:33:45 +0100
• 99ef63e019 Added simple (de-)serialization of (trained) nets Michael Mandl 2015-10-25 17:40:22 +0100
• 249bd22b67 Update README.md mandlm 2015-10-25 11:49:45 +0100
• 91427954c7 Added an icon to NeuroUI Michael Mandl 2015-10-25 10:00:48 +0100
• 2b84b8a054 Added a progress bar to NeuroUI Michael Mandl 2015-10-25 09:51:09 +0100
• 6943fc0116 Threaded learning and signaling in the Qt UI Michael Mandl 2015-10-24 18:03:07 +0200
• c9fb9b9fa8 Qt GUI initial import Michael Mandl 2015-10-24 14:44:33 +0200
• 4eb232b1e9 Working net, calculates the mean value of two inputs Michael Mandl 2015-10-23 22:16:12 +0200
• 6ed30e56c4 Finished the max-value net (2/3/1 neurons) with 10k learning iterations. No good. Michael Mandl 2015-10-22 22:09:35 +0200
• d4a22ecae7 Simple max(a, b) test-net without iterations Michael Mandl 2015-10-22 16:03:10 +0200
• f22e4221a1 Fixed feed-forward algo Michael Mandl 2015-10-22 16:02:27 +0200
• 3d30346f2d Implemented dynamic learning Michael Mandl 2015-10-18 22:05:18 +0200
• 6ef1f9657c Backprop seems to be working, yay Michael Mandl 2015-10-18 21:20:37 +0200
• a79abb5db1 First implementation of weight updates. Very slow rate of change in the output value. Michael Mandl 2015-10-17 22:05:27 +0200
• de06daaad3 Fixed gradient computation Michael Mandl 2015-10-17 21:02:10 +0200
• 26a51ce476 Hidden neuron gradients complete Michael Mandl 2015-10-16 23:23:27 +0200
• 370451c2e6 Calculation of hidden neuron gradients (partial) Michael Mandl 2015-10-16 22:59:04 +0200
• ea454b58c6 Added 64-bit configuration and support Michael Mandl 2015-10-15 22:37:13 +0200
• 953db0a17e Added README to vcxproj Michael Mandl 2015-10-15 22:30:53 +0200
• a7c0fb8b33 Create README.md mandlm 2015-10-15 22:19:48 +0200
• ce88f690cf More back-propagation code, calculation of output-neuron gradients Michael Mandl 2015-10-15 22:16:34 +0200
• 7ba16e9e9d Renamed a few things, started working on back-propagation Michael Mandl 2015-10-15 19:18:26 +0200
• 2f556d1b92 Added a (hacky) bias neuron Michael Mandl 2015-03-24 13:45:38 +0100
• e3a804242c Split up to different source files, entry-point for back propagation Michael Mandl 2015-03-23 21:58:30 +0100
• c8e08193f1 Initial forward-feeding net Michael Mandl 2015-03-23 19:28:29 +0100
• 341a88ee2e 🎊 Added .gitattributes & .gitignore files Michael Mandl 2015-03-23 19:28:10 +0100