Class: MachineLearningWorkbench::Optimizer::NaturalEvolutionStrategies::RNES

Inherits:
Base
  • Object
show all
Defined in:
lib/machine_learning_workbench/optimizer/natural_evolution_strategies/rnes.rb

Overview

Radial Natural Evolution Strategies

Direct Known Subclasses

FNES

Instance Attribute Summary collapse

Attributes inherited from Base

#best, #eye, #last_fits, #mu, #ndims, #obj_fn, #opt_type, #parallel_fit, #rescale_lrate, #rescale_popsize, #rng, #sigma

Instance Method Summary collapse

Methods inherited from Base

#cmaes_lrate, #cmaes_popsize, #cmaes_utilities, #initialize, #interface_methods, #lrate, #move_inds, #popsize, #sorted_inds, #standard_normal_sample, #standard_normal_samples, #utils

Constructor Details

This class inherits a constructor from MachineLearningWorkbench::Optimizer::NaturalEvolutionStrategies::Base

Instance Attribute Details

#variance ⇒ Object (readonly)

Returns the value of attribute variance.



6
7
8
 
# File 'lib/machine_learning_workbench/optimizer/natural_evolution_strategies/rnes.rb', line 6

def variance
  @variance
end

Instance Method Details

#convergence ⇒ Object

Estimate algorithm convergence based on variance



40
41
42
 
# File 'lib/machine_learning_workbench/optimizer/natural_evolution_strategies/rnes.rb', line 40

def convergence
  variance
end

#initialize_distribution(mu_init: 0, sigma_init: 1) ⇒ Object 



8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
 
# File 'lib/machine_learning_workbench/optimizer/natural_evolution_strategies/rnes.rb', line 8

def initialize_distribution mu_init: 0, sigma_init: 1
  @mu = case mu_init
    when Array
      raise ArgumentError unless mu_init.size == ndims
      NArray[mu_init]
    when Numeric
      NArray.new([1,ndims]).fill mu_init
    else
      raise ArgumentError, "Something is wrong with mu_init: #{mu_init}"
  end
  @variance = sigma_init
  @sigma = case sigma_init
  when Array
    raise ArgumentError "RNES uses single global variance"
  when Numeric
    NArray.new([ndims]).fill(variance).diag
  else
    raise ArgumentError, "Something is wrong with sigma_init: #{sigma_init}"
  end
end

#load(data) ⇒ Object

Raises:

  • (ArgumentError) 


48
49
50
51
52
53
 
# File 'lib/machine_learning_workbench/optimizer/natural_evolution_strategies/rnes.rb', line 48

def load data
  raise ArgumentError unless data.size == 2
  mu_ary, @variance = data
  @mu = mu_ary.to_na
  @sigma = eye * variance
end

#save ⇒ Object 



44
45
46
 
# File 'lib/machine_learning_workbench/optimizer/natural_evolution_strategies/rnes.rb', line 44

def save
  [mu.to_a, variance]
end

#train(picks: sorted_inds) ⇒ Object 



29
30
31
32
33
34
35
36
37
 
# File 'lib/machine_learning_workbench/optimizer/natural_evolution_strategies/rnes.rb', line 29

def train picks: sorted_inds
  g_mu = utils.dot(picks)
  # g_sigma = utils.dot(picks.row_norms**2 - ndims).first # back to scalar
  row_norms = NLinalg.norm picks, 2, axis:1
  g_sigma = utils.dot(row_norms**2 - ndims)[0] # back to scalar
  @mu += sigma.dot(g_mu.transpose).transpose * lrate
  @variance *= Math.exp(g_sigma * lrate / 2)
  @sigma = NArray.new([ndims]).fill(variance).diag
end