rm.operator.neural_net
Visibility: public
Uploaded 15-07-2016 by
Jan van Rijn
RapidMiner_6.4.0
0 runs
0 likes
downloaded by 0 people 0 issues
0 downvotes
, 0 total downloads
| Issue |
#Downvotes for this reason |
By |
|
Parameters
| decay | Indicates if the learning rate should be decreased during learningh | default: false |
| error_epsilon | The optimization is stopped if the training error gets below this epsilon value. | default: 1.0E-5 |
| hidden_layers | Describes the name and the size of all hidden layers. | |
| learning_rate | The learning rate determines by how much we change the weights at each step. May not be 0. | default: 0.3 |
| local_random_seed | Specifies the local random seed | default: 1992 |
| momentum | The momentum simply adds a fraction of the previous weight update to the current one (prevent local maxima and smoothes optimization directions). | default: 0.2 |
| normalize | Indicates if the input data should be normalized between -1 and +1 before learning (increases runtime but is in most cases necessary) | default: true |
| shuffle | Indicates if the input data should be shuffled before learning (increases memory usage but is recommended if data is sorted before) | default: true |
| training_cycles | The number of training cycles used for the neural network training. | default: 500 |
| use_local_random_seed | Indicates if a local random seed should be used. | default: false |
0
Runs
List all runs
Parameter: