eriklindernoren
|
2067d12c38
|
Cleaned up imports
|
2017-09-20 19:17:03 +02:00 |
|
eriklindernoren
|
e3a4d623d0
|
Code refinements. Batch iterator
|
2017-09-20 15:44:23 +02:00 |
|
eriklindernoren
|
d59c0fda41
|
mlfromscratch/deep_learning/
|
2017-09-19 00:14:57 +02:00 |
|
eriklindernoren
|
6e76164ed8
|
Padding layer fix.
|
2017-09-18 16:35:41 +02:00 |
|
eriklindernoren
|
d58485e573
|
Example folder
|
2017-09-18 16:30:09 +02:00 |
|
eriklindernoren
|
9f88013353
|
Recurrent Neural Network
|
2017-09-18 02:55:51 +02:00 |
|
eriklindernoren
|
98d247269e
|
Recurrent Neural Network
|
2017-09-18 02:40:39 +02:00 |
|
eriklindernoren
|
258fe6d1f4
|
Recurrent Neural Network
|
2017-09-18 02:11:16 +02:00 |
|
eriklindernoren
|
b7d00a1b52
|
Recurrent Neural Network
|
2017-09-18 01:40:40 +02:00 |
|
eriklindernoren
|
8f77243527
|
Fully-connected RNN layer
|
2017-09-18 01:04:51 +02:00 |
|
eriklindernoren
|
f4f2d6065b
|
Logistic Regression fix
|
2017-09-15 20:47:34 +02:00 |
|
eriklindernoren
|
445e711fab
|
Documentation
|
2017-09-15 19:18:45 +02:00 |
|
Qi, Bob
|
62830904df
|
use full module name for data_manipulation
|
2017-09-07 16:21:21 +08:00 |
|
eriklindernoren
|
b50b7a1aff
|
Loss functions
|
2017-08-21 10:53:38 +02:00 |
|
eriklindernoren
|
d17d4ddf06
|
ANN training speedup fix.
|
2017-08-21 10:36:53 +02:00 |
|
eriklindernoren
|
6efe4947c6
|
Deep Reinforcement Learning. OpenAI gym evaluation.
|
2017-08-18 14:52:33 +02:00 |
|
eriklindernoren
|
a9fbf0a04f
|
Neural Network: Model Summary
|
2017-08-18 10:24:49 +02:00 |
|
eriklindernoren
|
fef15b3052
|
Neural Network: AsciiTables for model summary
|
2017-08-17 10:50:02 +02:00 |
|
eriklindernoren
|
bc6075faee
|
Neural Network: Model summary. NeuralNetwork.fit() returns training error and validation error
|
2017-08-16 12:41:40 +02:00 |
|
eriklindernoren
|
2819c1711e
|
Batch Normalization: Initialization of running mean and running variance fix
|
2017-08-14 14:48:58 +02:00 |
|
eriklindernoren
|
72b6f6b08a
|
Easier construction of conv nets by automatic padding selection. Similarly to Keras. 'valid' or 'same' possible options.
|
2017-08-14 14:14:13 +02:00 |
|
eriklindernoren
|
ce7b704394
|
Deep Convolutional GAN
|
2017-08-13 19:18:50 +02:00 |
|
eriklindernoren
|
0f7ea66878
|
Refined GAN
|
2017-08-10 01:00:07 +02:00 |
|
eriklindernoren
|
d8012974f4
|
Refined batch norm impl.
|
2017-08-09 15:49:12 +02:00 |
|
eriklindernoren
|
d4549930fa
|
Batch Normalization layer implemented
|
2017-08-09 12:24:59 +02:00 |
|
eriklindernoren
|
bc31bc7497
|
Generative Adversarial Network impl.
|
2017-08-08 14:22:41 +02:00 |
|
eriklindernoren
|
663e4cc467
|
Constant padding and zero padding layers
|
2017-08-07 13:42:23 +02:00 |
|
eriklindernoren
|
46fea18e0e
|
Constant padding and zero padding layers
|
2017-08-07 13:37:01 +02:00 |
|
eriklindernoren
|
35d2d5e0db
|
Constant padding and zero padding layers
|
2017-08-07 13:34:49 +02:00 |
|
eriklindernoren
|
d7a9f3a5de
|
Constant padding and zero padding layers
|
2017-08-07 13:34:43 +02:00 |
|
eriklindernoren
|
5d17f2fc5e
|
One-hot encoding method renamed to 'to_categorical'
|
2017-08-06 18:17:58 +02:00 |
|
eriklindernoren
|
3b9bcce3aa
|
AveragePooling. Pooling parent class.
|
2017-08-06 17:14:22 +02:00 |
|
eriklindernoren
|
7a81dbae52
|
Max Pooling
|
2017-07-30 18:32:52 +02:00 |
|
eriklindernoren
|
0fc48fbd5f
|
Max Pooling
|
2017-07-30 18:28:12 +02:00 |
|
eriklindernoren
|
b6293b8575
|
Conv layer documention
|
2017-07-30 12:31:54 +02:00 |
|
eriklindernoren
|
ebf99cd538
|
Convolutional Neural Network
|
2017-07-30 05:50:56 +02:00 |
|
eriklindernoren
|
74aba24c32
|
Deep nets: Moved layers to 'mlfromscratch/utils/layers.py'. Keras Sequential-like network construction.
|
2017-07-28 21:53:21 +02:00 |
|
eriklindernoren
|
6af5f087b4
|
Deep nets: Moved layers to 'mlfromscratch/utils/layers.py'. Keras Sequential-like network construction.
|
2017-07-28 21:42:30 +02:00 |
|
eriklindernoren
|
980e9621ed
|
Dropout layer impl.
|
2017-06-18 17:13:29 +02:00 |
|
eriklindernoren
|
b9766b5af1
|
Simple matplotlib wrapper with dim. reduction implemented in utils/misc.py.
|
2017-06-17 23:16:38 +02:00 |
|
eriklindernoren
|
4eb90be16b
|
Multilayer Perceptron: Support for implemented optimizers. Softmax activation function and cross entropy loss implemented.
|
2017-06-12 00:35:40 +02:00 |
|
eriklindernoren
|
ea2253f314
|
Activation function: Scaled Exponential Linear Unit
|
2017-06-10 16:51:43 +02:00 |
|
eriklindernoren
|
5890fb9da6
|
Activation function: Scaled Exponential Linear Unit
|
2017-06-10 16:36:42 +02:00 |
|
eriklindernoren
|
060b486974
|
Apriori example output
|
2017-06-09 16:48:31 +02:00 |
|
eriklindernoren
|
8fc3b4bfd2
|
PyPI
|
2017-06-08 02:15:16 +02:00 |
|