ds.monitored_traindbm: Fine-Tuning of a DBM

In stefan-m-lenz/dsBoltzmannMachinesClient: Client for dsBoltzmannMachines

Description

This functions performs monitored fine-tuning of a given DBM model. For the complete training, including the pre-training, see ds.monitored_fitdbm. During the training, monitoring data is collected by default. The monitoring data is returned to the user. The trained model is stored on the server side (see parameter newobj).

Usage

ds.monitored_traindbm(
  datasources,
  dbm = "dbm",
  newobj = "dbm",
  data = "D",
  monitoring = "logproblowerbound",
  monitoringdata = data,
  epochs = NULL,
  nparticles = NULL,
  learningrate = NULL,
  learningrates = NULL
)

Arguments

dbm	The name of DBM model that is to be fine-tuned. Defaults to "dbm".
newobj	The name of the variable to store the new DBM model. Defaults to "dbm", such that the previous model is overwritten.
monitoring	Name(s) for monitoring options used for DBM training. For possible options, see ds.monitored_fitdbm
monitoringdata	A vector of names of server-side data sets that are to be used for monitoring
epochs	Number of training epochs for fine-tuning, defaults to 10
nparticles	Number of particles used for sampling during fine-tuning of the DBM, defaults to 100
learningrate	Learning rate for fine-tuning, decaying by default decaying with the number of epochs, starting with the given value for the learningrate. By default, the learning rate decreases with the factor 11 / (10 + epoch).
learningrates	A vector of learning rates for each epoch of fine-tuning

Details

If the option dsBoltzmannMachines.shareModels is set to TRUE by an administrator at the server side, the model itself is returned in addition.

stefan-m-lenz/dsBoltzmannMachinesClient documentation built on May 2, 2021, 10:46 a.m.