|
|
a3be9daf02
|
Added LeakyReLu (parameter is currently hardcoded)
|
2022-05-25 17:28:03 +02:00 |
|
|
|
f1ca0a9e54
|
Changed initialization of weights to gaussian distributed values and biases to 1 (this helps (Leaky)ReLu and doesn't seem hurt tanh yet)
|
2022-05-25 17:26:55 +02:00 |
|
|
|
281b42b0fb
|
Added XChart library to plot graphs
|
2022-05-25 15:34:28 +02:00 |
|
|
|
adfd701817
|
Added support for creating matrices with Gaussian distributed values
|
2022-05-25 15:33:17 +02:00 |
|
|
|
ffcf9fa975
|
Fixed typo
|
2022-05-25 02:41:17 +02:00 |
|
|
|
e02e79308f
|
Extracted the gradient descent example into a separate class
|
2022-05-25 02:40:58 +02:00 |
|
|
|
db0481e9cf
|
Added gradient descent for vector-valued functions
|
2022-05-25 02:39:18 +02:00 |
|
|
|
1c66f1b72f
|
The inputSize of each layer does not need to be specified anymore
|
2022-05-25 00:41:38 +02:00 |
|
|
|
95501bf4b1
|
Moved support for adding neurons to FCLayer because BlankLayer is pretty much the same as FCLayer and currently useless
|
2022-05-25 00:17:33 +02:00 |
|
|
|
4766ea0ad9
|
Added additional XOR example with later added neurons
|
2022-05-23 19:10:43 +02:00 |
|
|
|
8c82838c54
|
Added support for adding new neurons
|
2022-05-23 19:09:26 +02:00 |
|
|
|
7738781bb5
|
Added some utility functions (not yet needed)
|
2022-05-23 19:08:50 +02:00 |
|
|
|
7e80e5bc94
|
Deleted .idea/workspace.xml and added it to .gitignore
|
2022-05-22 22:00:26 +02:00 |
|
|
|
d0e48cd8fa
|
Added README.md
|
2022-05-22 21:08:45 +02:00 |
|
|
|
fc1d803c99
|
Added ReLu activation function
|
2022-05-22 01:47:39 +02:00 |
|
|
|
69a6c8df29
|
Added MAE loss function
|
2022-05-22 01:42:41 +02:00 |
|
|
|
8376c00470
|
Increased range of inital weights and biases to improve results
|
2022-05-22 01:17:06 +02:00 |
|
|
|
5aa9313776
|
Added support for different activation and loss functions
|
2022-05-22 01:11:38 +02:00 |
|
|
|
645a8baf2c
|
Commented out not yet working code
|
2022-05-21 23:32:07 +02:00 |
|
|
|
a431e08264
|
Initial commit
|
2022-05-21 23:15:51 +02:00 |
|
