MOAoptions: Get and set options for models build with MOA.

Description Usage Arguments Value Examples

View source: R/MOAmodeloptions.R

Description

Get and set options for models build with MOA.

Usage

1
MOAoptions(model, ...)

Arguments

model

character string with a model or an object of class MOA_model. E.g. HoeffdingTree, DecisionStump, NaiveBayes, HoeffdingOptionTree, ... The list of known models can be obtained by typing RMOA:::.moaknownmodels. See the examples.

...

other parameters specifying the MOA modelling options of each model. See the examples.

Value

An object of class MOAmodelOptions.
This is a list with elements:

  1. model: The name of the model

  2. moamodelname: The purpose of the model known by MOA (getPurposeString)

  3. javaObj: a java reference of MOA options

  4. options: a list with options of the MOA model. Each list element contains the Name of the option, the Purpose of the option and the current Value

See the examples.

Examples

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
control <- MOAoptions(model = "HoeffdingTree")
control
MOAoptions(model = "HoeffdingTree", leafprediction = "MC",
   removePoorAtts = TRUE, binarySplits = TRUE, tieThreshold = 0.20)

## Other models known by RMOA
RMOA:::.moaknownmodels

## Classification Trees
MOAoptions(model = "AdaHoeffdingOptionTree")
MOAoptions(model = "ASHoeffdingTree")
MOAoptions(model = "DecisionStump")
MOAoptions(model = "HoeffdingAdaptiveTree")
MOAoptions(model = "HoeffdingOptionTree")
MOAoptions(model = "HoeffdingTree")
MOAoptions(model = "LimAttHoeffdingTree")
MOAoptions(model = "RandomHoeffdingTree")
## Classification using Bayes rule
MOAoptions(model = "NaiveBayes")
MOAoptions(model = "NaiveBayesMultinomial")
## Classification using Active learning
MOAoptions(model = "ActiveClassifier")
## Classification using Ensemble learning
MOAoptions(model = "AccuracyUpdatedEnsemble")
MOAoptions(model = "AccuracyWeightedEnsemble")
MOAoptions(model = "ADACC")
MOAoptions(model = "DACC")
MOAoptions(model = "LeveragingBag")
MOAoptions(model = "OCBoost")
MOAoptions(model = "OnlineAccuracyUpdatedEnsemble")
MOAoptions(model = "OzaBag")
MOAoptions(model = "OzaBagAdwin")
MOAoptions(model = "OzaBagASHT")
MOAoptions(model = "OzaBoost")
MOAoptions(model = "OzaBoostAdwin")
MOAoptions(model = "TemporallyAugmentedClassifier")
MOAoptions(model = "WeightedMajorityAlgorithm")

## Regressions
MOAoptions(model = "AMRulesRegressor")
MOAoptions(model = "FadingTargetMean")
MOAoptions(model = "FIMTDD")
MOAoptions(model = "ORTO")
MOAoptions(model = "Perceptron")
MOAoptions(model = "SGD")
MOAoptions(model = "TargetMean")

Example output

Loading required package: RMOAjars
OpenJDK 64-Bit Server VM warning: Can't detect initial thread stack location - find_vma failed
Loading required package: rJava
HoeffdingTree modelling options: 
MOA model name: Hoeffding Tree or VFDT.
  - maxByteSize: 33554432   (Maximum memory consumed by the tree.)
  - numericEstimator: GaussianNumericAttributeClassObserver   (Numeric estimator to use.)
  - nominalEstimator: NominalAttributeClassObserver   (Nominal estimator to use.)
  - memoryEstimatePeriod: 1000000   (How many instances between memory consumption checks.)
  - gracePeriod: 200   (The number of instances a leaf should observe between split attempts.)
  - splitCriterion: InfoGainSplitCriterion   (Split criterion to use.)
  - splitConfidence: 1e-07   (The allowable error in split decision, values closer to 0 will take longer to decide.)
  - tieThreshold: 0.05   (Threshold below which a split will be forced to break ties.)
  - binarySplits: false   (Only allow binary splits.)
  - stopMemManagement: false   (Stop growing as soon as memory limit is hit.)
  - removePoorAtts: false   (Disable poor attributes.)
  - noPrePrune: false   (Disable pre-pruning.)
  - leafprediction: NBAdaptive   (Leaf prediction to use.)
  - nbThreshold: 0   (The number of instances a leaf should observe before permitting Naive Bayes.)
HoeffdingTree modelling options: 
MOA model name: Hoeffding Tree or VFDT.
  - maxByteSize: 33554432   (Maximum memory consumed by the tree.)
  - numericEstimator: GaussianNumericAttributeClassObserver   (Numeric estimator to use.)
  - nominalEstimator: NominalAttributeClassObserver   (Nominal estimator to use.)
  - memoryEstimatePeriod: 1000000   (How many instances between memory consumption checks.)
  - gracePeriod: 200   (The number of instances a leaf should observe between split attempts.)
  - splitCriterion: InfoGainSplitCriterion   (Split criterion to use.)
  - splitConfidence: 1e-07   (The allowable error in split decision, values closer to 0 will take longer to decide.)
  - tieThreshold: 0.2   (Threshold below which a split will be forced to break ties.)
  - binarySplits: true   (Only allow binary splits.)
  - stopMemManagement: false   (Stop growing as soon as memory limit is hit.)
  - removePoorAtts: true   (Disable poor attributes.)
  - noPrePrune: false   (Disable pre-pruning.)
  - leafprediction: MC   (Leaf prediction to use.)
  - nbThreshold: 0   (The number of instances a leaf should observe before permitting Naive Bayes.)
 [1] "AdaHoeffdingOptionTree"        "ASHoeffdingTree"              
 [3] "DecisionStump"                 "HoeffdingAdaptiveTree"        
 [5] "HoeffdingOptionTree"           "HoeffdingTree"                
 [7] "LimAttHoeffdingTree"           "RandomHoeffdingTree"          
 [9] "NaiveBayes"                    "NaiveBayesMultinomial"        
[11] "ActiveClassifier"              "AccuracyUpdatedEnsemble"      
[13] "AccuracyWeightedEnsemble"      "ADACC"                        
[15] "DACC"                          "LeveragingBag"                
[17] "LimAttClassifier"              "OCBoost"                      
[19] "OnlineAccuracyUpdatedEnsemble" "OzaBag"                       
[21] "OzaBagAdwin"                   "OzaBagASHT"                   
[23] "OzaBoost"                      "OzaBoostAdwin"                
[25] "TemporallyAugmentedClassifier" "WeightedMajorityAlgorithm"    
[27] "AMRulesRegressor"              "FadingTargetMean"             
[29] "FIMTDD"                        "ORTO"                         
[31] "Perceptron"                    "SGD"                          
[33] "TargetMean"                   
AdaHoeffdingOptionTree modelling options: 
MOA model name: Adaptive decision option tree for streaming data with adaptive Naive Bayes classification at leaves.
  - maxOptionPaths: 5   (Maximum number of option paths per node.)
  - maxByteSize: 33554432   (Maximum memory consumed by the tree.)
  - numericEstimator: GaussianNumericAttributeClassObserver   (Numeric estimator to use.)
  - nominalEstimator: NominalAttributeClassObserver   (Nominal estimator to use.)
  - memoryEstimatePeriod: 1000000   (How many instances between memory consumption checks.)
  - gracePeriod: 200   (The number of instances a leaf should observe between split attempts.)
  - splitCriterion: InfoGainSplitCriterion   (Split criterion to use.)
  - splitConfidence: 1e-07   (The allowable error in split decision, values closer to 0 will take longer to decide.)
  - secondarySplitConfidence: 0.1   (The allowable error in secondary split decisions, values closer to 0 will take longer to decide.)
  - tieThreshold: 0.05   (Threshold below which a split will be forced to break ties.)
  - binarySplits: false   (Only allow binary splits.)
  - removePoorAtts: false   (Disable poor attributes.)
  - noPrePrune: false   (Disable pre-pruning.)

  - memStrategy: 2   (Memory strategy to use.)
  - leafprediction: NBAdaptive   (Leaf prediction to use.)
  - nbThreshold: 0   (The number of instances a leaf should observe before permitting Naive Bayes.)
ASHoeffdingTree modelling options: 
MOA model name: Adaptive Size Hoeffding Tree used in Bagging using trees of different size.
  - maxByteSize: 33554432   (Maximum memory consumed by the tree.)
  - numericEstimator: GaussianNumericAttributeClassObserver   (Numeric estimator to use.)
  - nominalEstimator: NominalAttributeClassObserver   (Nominal estimator to use.)
  - memoryEstimatePeriod: 1000000   (How many instances between memory consumption checks.)
  - gracePeriod: 200   (The number of instances a leaf should observe between split attempts.)
  - splitCriterion: InfoGainSplitCriterion   (Split criterion to use.)
  - splitConfidence: 1e-07   (The allowable error in split decision, values closer to 0 will take longer to decide.)
  - tieThreshold: 0.05   (Threshold below which a split will be forced to break ties.)
  - binarySplits: false   (Only allow binary splits.)
  - stopMemManagement: false   (Stop growing as soon as memory limit is hit.)
  - removePoorAtts: false   (Disable poor attributes.)
  - noPrePrune: false   (Disable pre-pruning.)
  - leafprediction: NBAdaptive   (Leaf prediction to use.)
  - nbThreshold: 0   (The number of instances a leaf should observe before permitting Naive Bayes.)
DecisionStump modelling options: 
MOA model name: Decision trees of one level.
  - gracePeriod: 1000   (The number of instances to observe between model changes.)
  - binarySplits: false   (Only allow binary splits.)
  - splitCriterion: InfoGainSplitCriterion   (Split criterion to use.)
HoeffdingAdaptiveTree modelling options: 
MOA model name: Hoeffding Adaptive Tree for evolving data streams that uses ADWIN to replace branches for new ones.
  - maxByteSize: 33554432   (Maximum memory consumed by the tree.)
  - numericEstimator: GaussianNumericAttributeClassObserver   (Numeric estimator to use.)
  - nominalEstimator: NominalAttributeClassObserver   (Nominal estimator to use.)
  - memoryEstimatePeriod: 1000000   (How many instances between memory consumption checks.)
  - gracePeriod: 200   (The number of instances a leaf should observe between split attempts.)
  - splitCriterion: InfoGainSplitCriterion   (Split criterion to use.)
  - splitConfidence: 1e-07   (The allowable error in split decision, values closer to 0 will take longer to decide.)
  - tieThreshold: 0.05   (Threshold below which a split will be forced to break ties.)
  - binarySplits: false   (Only allow binary splits.)
  - stopMemManagement: false   (Stop growing as soon as memory limit is hit.)
  - removePoorAtts: false   (Disable poor attributes.)
  - noPrePrune: false   (Disable pre-pruning.)
  - leafprediction: NBAdaptive   (Leaf prediction to use.)
  - nbThreshold: 0   (The number of instances a leaf should observe before permitting Naive Bayes.)
HoeffdingOptionTree modelling options: 
MOA model name: Hoeffding Option Tree: single tree that represents multiple trees.
  - maxOptionPaths: 5   (Maximum number of option paths per node.)
  - maxByteSize: 33554432   (Maximum memory consumed by the tree.)
  - numericEstimator: GaussianNumericAttributeClassObserver   (Numeric estimator to use.)
  - nominalEstimator: NominalAttributeClassObserver   (Nominal estimator to use.)
  - memoryEstimatePeriod: 1000000   (How many instances between memory consumption checks.)
  - gracePeriod: 200   (The number of instances a leaf should observe between split attempts.)
  - splitCriterion: InfoGainSplitCriterion   (Split criterion to use.)
  - splitConfidence: 1e-07   (The allowable error in split decision, values closer to 0 will take longer to decide.)
  - secondarySplitConfidence: 0.1   (The allowable error in secondary split decisions, values closer to 0 will take longer to decide.)
  - tieThreshold: 0.05   (Threshold below which a split will be forced to break ties.)
  - binarySplits: false   (Only allow binary splits.)
  - removePoorAtts: false   (Disable poor attributes.)
  - noPrePrune: false   (Disable pre-pruning.)

  - memStrategy: 2   (Memory strategy to use.)
  - leafprediction: NBAdaptive   (Leaf prediction to use.)
  - nbThreshold: 0   (The number of instances a leaf should observe before permitting Naive Bayes.)
HoeffdingTree modelling options: 
MOA model name: Hoeffding Tree or VFDT.
  - maxByteSize: 33554432   (Maximum memory consumed by the tree.)
  - numericEstimator: GaussianNumericAttributeClassObserver   (Numeric estimator to use.)
  - nominalEstimator: NominalAttributeClassObserver   (Nominal estimator to use.)
  - memoryEstimatePeriod: 1000000   (How many instances between memory consumption checks.)
  - gracePeriod: 200   (The number of instances a leaf should observe between split attempts.)
  - splitCriterion: InfoGainSplitCriterion   (Split criterion to use.)
  - splitConfidence: 1e-07   (The allowable error in split decision, values closer to 0 will take longer to decide.)
  - tieThreshold: 0.05   (Threshold below which a split will be forced to break ties.)
  - binarySplits: false   (Only allow binary splits.)
  - stopMemManagement: false   (Stop growing as soon as memory limit is hit.)
  - removePoorAtts: false   (Disable poor attributes.)
  - noPrePrune: false   (Disable pre-pruning.)
  - leafprediction: NBAdaptive   (Leaf prediction to use.)
  - nbThreshold: 0   (The number of instances a leaf should observe before permitting Naive Bayes.)
LimAttHoeffdingTree modelling options: 
MOA model name: Hoeffding decision trees with a restricted number of attributes for data streams.
  - maxByteSize: 33554432   (Maximum memory consumed by the tree.)
  - numericEstimator: GaussianNumericAttributeClassObserver   (Numeric estimator to use.)
  - nominalEstimator: NominalAttributeClassObserver   (Nominal estimator to use.)
  - memoryEstimatePeriod: 1000000   (How many instances between memory consumption checks.)
  - gracePeriod: 200   (The number of instances a leaf should observe between split attempts.)
  - splitCriterion: InfoGainSplitCriterion   (Split criterion to use.)
  - splitConfidence: 1e-07   (The allowable error in split decision, values closer to 0 will take longer to decide.)
  - tieThreshold: 0.05   (Threshold below which a split will be forced to break ties.)
  - binarySplits: false   (Only allow binary splits.)
  - stopMemManagement: false   (Stop growing as soon as memory limit is hit.)
  - noPrePrune: false   (Disable pre-pruning.)
  - leafprediction: NBAdaptive   (Leaf prediction to use.)
  - nbThreshold: 0   (The number of instances a leaf should observe before permitting Naive Bayes.)
  - randomSeed: 1   (Seed for random behaviour of the classifier.)
RandomHoeffdingTree modelling options: 
MOA model name: Random decision trees for data streams.
  - maxByteSize: 33554432   (Maximum memory consumed by the tree.)
  - numericEstimator: GaussianNumericAttributeClassObserver   (Numeric estimator to use.)
  - nominalEstimator: NominalAttributeClassObserver   (Nominal estimator to use.)
  - memoryEstimatePeriod: 1000000   (How many instances between memory consumption checks.)
  - gracePeriod: 200   (The number of instances a leaf should observe between split attempts.)
  - splitCriterion: InfoGainSplitCriterion   (Split criterion to use.)
  - splitConfidence: 1e-07   (The allowable error in split decision, values closer to 0 will take longer to decide.)
  - tieThreshold: 0.05   (Threshold below which a split will be forced to break ties.)
  - binarySplits: false   (Only allow binary splits.)
  - stopMemManagement: false   (Stop growing as soon as memory limit is hit.)
  - noPrePrune: false   (Disable pre-pruning.)
  - leafprediction: NBAdaptive   (Leaf prediction to use.)
  - nbThreshold: 0   (The number of instances a leaf should observe before permitting Naive Bayes.)
  - randomSeed: 1   (Seed for random behaviour of the classifier.)
NaiveBayes modelling options: 
MOA model name: Naive Bayes classifier: performs classic bayesian prediction while making naive assumption that all inputs are independent.
NaiveBayesMultinomial modelling options: 
MOA model name: AMultinomial Naive Bayes classifier: performs classic bayesian prediction while making naive assumption that all inputs are independent.
  - laplaceCorrection: 1   (Laplace correction factor.)
ActiveClassifier modelling options: 
MOA model name: Active learning classifier for evolving data streams
  - baseLearner: drift.SingleClassifierDrift   (Classifier to train.)
  - activeLearningStrategy: Random   (Active Learning Strategy to use.)
  - budget: 0.1   (Budget to use.)
  - fixedThreshold: 0.9   (Fixed threshold.)
  - step: 0.01   (Floating budget step.)
  - numInstancesInit: 0   (Number of instances at beginning without active learning.)
  - randomSeed: 1   (Seed for random behaviour of the classifier.)
AccuracyUpdatedEnsemble modelling options: 
MOA model name: MOA Classifier: moa.classifiers.meta.AccuracyUpdatedEnsemble
  - learner: trees.HoeffdingTree -e 2000000 -g 100 -c 0.01   (Classifier to train.)
  - memberCount: 10   (The maximum number of classifiers in an ensemble.)
  - chunkSize: 500   (The chunk size used for classifier creation and evaluation.)
  - maxByteSize: 33554432   (Maximum memory consumed by ensemble.)
AccuracyWeightedEnsemble modelling options: 
MOA model name: Accuracy Weighted Ensemble classifier as proposed by Wang et al. in 'Mining concept-drifting data streams using ensemble classifiers', KDD 2003
  - learner: trees.HoeffdingTree -e 1000 -g 100 -c 0.01 -l NB   (Classifier to train.)
  - memberCount: 15   (The maximum number of classifier in an ensemble.)
  - storedCount: 30   (The maximum number of classifiers to store and choose from when creating an ensemble.)
  - chunkSize: 500   (The chunk size used for classifier creation and evaluation.)
  - numFolds: 10   (Number of cross-validation folds for candidate classifier testing.)
ADACC modelling options: 
MOA model name: Anticipative and Dynamic Adaptation to Concept Changes for data streams.
  - tau: 100   (The size of the evaluation window for the meta-learning.)
  - StabThr: 0.8   (The threshold for stability)
  - CeThr: 0.7   (The threshold for concept equivalence)
  - baseLearner: bayes.NaiveBayes   (Classifier to train.)
  - memberCount: 20   (The maximum number of classifier in an ensemble.)
  - maturity: 20   (The maturity age.)
  - eval: 20   (The size of the evaluation window.)
  - cmb: MAX   (The combination function)
  - randomSeed: 1   (Seed for random behaviour of the classifier.)
DACC modelling options: 
MOA model name: Dynamic Adaptation to Concept Changes for data streams.
  - baseLearner: bayes.NaiveBayes   (Classifier to train.)
  - memberCount: 20   (The maximum number of classifier in an ensemble.)
  - maturity: 20   (The maturity age.)
  - eval: 20   (The size of the evaluation window.)
  - cmb: MAX   (The combination function)
  - randomSeed: 1   (Seed for random behaviour of the classifier.)
LeveragingBag modelling options: 
MOA model name: Leveraging Bagging for evolving data streams using ADWIN.
  - baseLearner: trees.HoeffdingTree   (Classifier to train.)
  - ensembleSize: 10   (The number of models in the bag.)
  - weightShrink: 6   (The number to use to compute the weight of new instances.)
  - deltaAdwin: 0.002   (Delta of Adwin change detection)
  - outputCodes: false   (Use Output Codes to use binary classifiers.)
  - leveraginBagAlgorithm: LeveragingBag   (Leveraging Bagging to use.)
  - randomSeed: 1   (Seed for random behaviour of the classifier.)
OCBoost modelling options: 
MOA model name: Online Coordinate boosting for two classes evolving data streams.
  - baseLearner: trees.HoeffdingTree   (Classifier to train.)
  - ensembleSize: 10   (The number of models to boost.)
  - smoothingParameter: 0.5   (Smoothing parameter.)
  - randomSeed: 1   (Seed for random behaviour of the classifier.)
OnlineAccuracyUpdatedEnsemble modelling options: 
MOA model name: MOA Classifier: moa.classifiers.meta.OnlineAccuracyUpdatedEnsemble
  - learner: trees.HoeffdingTree -e 2000000 -g 100 -c 0.01   (Classifier to train.)
  - memberCount: 10   (The maximum number of classifiers in an ensemble.)
  - windowSize: 500   (The window size used for classifier creation and evaluation.)
  - maxByteSize: 33554432   (Maximum memory consumed by ensemble.)
  - verbose: false   (When checked the algorithm outputs additional information about component classifier weights.)
  - linearFunction: false   (When checked the algorithm uses a linear weighting function.)
OzaBag modelling options: 
MOA model name: Incremental on-line bagging of Oza and Russell.
  - baseLearner: trees.HoeffdingTree   (Classifier to train.)
  - ensembleSize: 10   (The number of models in the bag.)
  - randomSeed: 1   (Seed for random behaviour of the classifier.)
OzaBagAdwin modelling options: 
MOA model name: Bagging for evolving data streams using ADWIN.
  - baseLearner: trees.HoeffdingTree   (Classifier to train.)
  - ensembleSize: 10   (The number of models in the bag.)
  - randomSeed: 1   (Seed for random behaviour of the classifier.)
OzaBagASHT modelling options: 
MOA model name: Bagging using trees of different size.
  - firstClassifierSize: 1   (The size of first classifier in the bag.)
  - useWeight: false   (Enable weight classifiers.)
  - resetTrees: false   (Reset trees when size is higher than the max.)
  - baseLearner: trees.HoeffdingTree   (Classifier to train.)
  - ensembleSize: 10   (The number of models in the bag.)
  - randomSeed: 1   (Seed for random behaviour of the classifier.)
OzaBoost modelling options: 
MOA model name: Incremental on-line boosting of Oza and Russell.
  - baseLearner: trees.HoeffdingTree   (Classifier to train.)
  - ensembleSize: 10   (The number of models to boost.)
  - pureBoost: false   (Boost with weights only; no poisson.)
  - randomSeed: 1   (Seed for random behaviour of the classifier.)
OzaBoostAdwin modelling options: 
MOA model name: Boosting for evolving data streams using ADWIN.
  - baseLearner: trees.HoeffdingTree   (Classifier to train.)
  - ensembleSize: 10   (The number of models to boost.)
  - pureBoost: false   (Boost with weights only; no poisson.)
  - deltaAdwin: 0.002   (Delta of Adwin change detection)
  - outputCodes: false   (Use Output Codes to use binary classifiers.)
  - same: false   (Use Samme Algorithm.)
  - randomSeed: 1   (Seed for random behaviour of the classifier.)
TemporallyAugmentedClassifier modelling options: 
MOA model name: Add some old labels to every instance
  - baseLearner: trees.HoeffdingTree   (Classifier to train.)
  - numOldLabels: 1   (The number of old labels to add to each example.)
  - labelDelay: false   (Labels arrive with Delay. Use predictions instead of true Labels.)
WeightedMajorityAlgorithm modelling options: 
MOA model name: Weighted majority algorithm for data streams.
  - learners: trees.HoeffdingTree -l MC,trees.HoeffdingTree -l NB,trees.HoeffdingTree,bayes.NaiveBayes   (The learners to combine.)
  - beta: 0.9   (Factor to punish mistakes by.)
  - gamma: 0.01   (Minimum fraction of weight per model.)
  - prune: false   (Prune poorly performing models from ensemble.)
AMRulesRegressor modelling options: 
MOA model name: MOA Classifier: moa.classifiers.rules.AMRulesRegressor
  - learningRatio_Decay_set_constant: false   (Learning Ratio Decay in Perceptron set to be constant. (The next parameter).)
  - learningRatio: 0.025   (Constante Learning Ratio to use for training the Perceptrons in the leaves.)
  - predictionFunctionOption: Adaptative   (The prediction function to use.)
  - votingType: InverseErrorWeightedVote   (Voting Type.)
  - splitConfidence: 1e-07   (Hoeffding Bound Parameter. The allowable error in split decision, values closer to 0 will take longer to decide.)
  - tieThreshold: 0.05   (Hoeffding Bound Parameter. Threshold below which a split will be forced to break ties.)
  - gracePeriod: 200   (Hoeffding Bound Parameter. The number of instances a leaf should observe between split attempts.)
  - DoNotDetectChanges: false   (Drift Detection. Page-Hinkley.)
  - pageHinckleyAlpha: 0.005   (The alpha value to use in the Page Hinckley change detection tests.)
  - pageHinckleyThreshold: 35   (The threshold value (Lambda) to be used in the Page Hinckley change detection tests.)
  - noAnomalyDetection: false   (Disable anomaly Detection.)
  - multivariateAnomalyProbabilityThresholdd: 0.99   (Multivariate anomaly threshold value.)
  - univariateAnomalyprobabilityThreshold: 0.1   (Univariate anomaly threshold value.)
  - anomalyThreshold: 30   (The threshold value of anomalies to be used in the anomaly detection.)
  - setUnorderedRulesOn: false   (unorderedRules.)
  - verbosity: 1   (Output Verbosity Control Level. 1 (Less) to 5 (More))
  - numericObserver: FIMTDDNumericAttributeClassLimitObserver   (Numeric observer.)
  - randomSeed: 1   (Seed for random behaviour of the classifier.)
FadingTargetMean modelling options: 
MOA model name: MOA Classifier: moa.classifiers.rules.functions.FadingTargetMean
  - fadingFactor: 0.99   (Fading factor for the FadingTargetMean accumulated error)
  - fadingErrorFactor: 0.99   (Fading error factor for the TargetMean accumulated error)
FIMTDD modelling options: 
MOA model name: Implementation of the FIMT-DD tree as described by Ikonomovska et al.
  - PageHinckleyAlpha: 0.005   (The alpha value to use in the Page Hinckley change detection tests.)
  - PageHinckleyThreshold: 50   (The threshold value to be used in the Page Hinckley change detection tests.)
  - AlternateTreeFadingFactor: 0.995   (The fading factor to use when deciding if an alternate tree should replace an original.)
  - AlternateTreeTMin: 150   (The Tmin value to use when deciding if an alternate tree should replace an original.)
  - AlternateTreeTime: 1500   (The 'time' (in terms of number of instances) value to use when deciding if an alternate tree should be discarded.)
  - learningRatio: 0.01   (Learning ratio to use for training the Perceptrons in the leaves.)
  - learningRatio_Decay_or_Const: false   (learning Ratio Decay or const parameter.)
  - maxByteSize: 33554432   (Maximum memory consumed by the tree.)
  - numericEstimator: FIMTDDNumericAttributeClassObserver   (Numeric estimator to use.)
  - nominalEstimator: NominalAttributeClassObserver   (Nominal estimator to use.)
  - memoryEstimatePeriod: 1000000   (How many instances between memory consumption checks.)
  - gracePeriod: 200   (The number of instances a leaf should observe between split attempts.)
  - splitCriterion: SDRSplitCriterion   (Split criterion to use.)
  - splitConfidence: 1e-07   (The allowable error in split decision, values closer to 0 will take longer to decide.)
  - tieThreshold: 0.05   (Threshold below which a split will be forced to break ties.)
  - binarySplits: false   (Only allow binary splits.)
  - stopMemManagement: false   (Stop growing as soon as memory limit is hit.)
  - removePoorAtts: false   (Disable poor attributes.)
  - noPrePrune: false   (Disable pre-pruning.)
  - leafprediction: NBAdaptive   (Leaf prediction to use.)
  - nbThreshold: 0   (The number of instances a leaf should observe before permitting Naive Bayes.)
ORTO modelling options: 
MOA model name: Implementation of the ORTO tree as described by Ikonomovska et al.
  - PageHinckleyAlpha: 0.005   (The alpha value to use in the Page Hinckley change detection tests.)
  - PageHinckleyThreshold: 50   (The threshold value to be used in the Page Hinckley change detection tests.)
  - AlternateTreeFadingFactor: 0.995   (The fading factor to use when deciding if an alternate tree should replace an original.)
  - AlternateTreeTMin: 150   (The Tmin value to use when deciding if an alternate tree should replace an original.)
  - AlternateTreeTime: 1500   (The 'time' (in terms of number of instances) value to use when deciding if an alternate tree should be discarded.)
  - LearningRatio: 0.01   (Learning ratio to use for training the Perceptrons in the leaves.)
  - LearningRatioDecayOrConst: false   (learning Ratio Decay or const parameter.)
  - MaxTrees: 10   (The maximum number of trees contained in the option tree.)
  - MaxOptionLevel: 10   (The maximal depth at which option nodes can be created.)
  - OptionDecayFactor: 0.9   (The option decay factor that determines how many options can be selected at a given level.)
  - splitCriterion: VarianceReductionSplitCriterion   (Split criterion to use.)
  - numericEstimator: FIMTDDNumericAttributeClassObserver   (Numeric estimator to use.)
  - gracePeriod: 200   (The number of instances a leaf should observe between split attempts.)
  - splitConfidence: 1e-07   (The allowable error in split decision, values closer to 0 will take longer to decide.)
  - tieThreshold: 0.05   (Threshold below which a split will be forced to break ties.)
  - removePoorAtts: false   (Disable poor attributes.)
  - OptionNodeAggregation: average   (The aggregation method used to combine predictions in option nodes.)
  - OptionFadingFactor: 0.9995   (The fading factor used for comparing subtrees of an option node.)
  - randomSeed: 1   (Seed for random behaviour of the classifier.)
Perceptron modelling options: 
MOA model name: MOA Classifier: moa.classifiers.rules.functions.Perceptron
  - learningRatio_Decay_set_constant: false   (Learning Ratio Decay in Perceptron set to be constant. (The next parameter).)
  - learningRatio: 0.01   (Constante Learning Ratio to use for training the Perceptrons in the leaves.)
  - learningRateDecay: 0.001   ( Learning Rate decay to use for training the Perceptron.)
  - fadingFactor: 0.99   (Fading factor for the Perceptron accumulated error)
  - randomSeed: 1   (Seed for random behaviour of the classifier.)
SGD modelling options: 
MOA model name: AStochastic gradient descent for learning various linear models (binary class SVM, binary class logistic regression and linear regression).
  - lambdaRegularization: 1e-04   (Lambda regularization parameter .)
  - learningRate: 1e-04   (Learning rate parameter.)
  - lossFunction: HINGE   (The loss function to use.)
TargetMean modelling options: 
MOA model name: MOA Classifier: moa.classifiers.rules.functions.TargetMean
  - fadingErrorFactor: 0.99   (Fading error factor for the TargetMean accumulated error)
Warning message:
system call failed: Cannot allocate memory 

RMOA documentation built on May 29, 2017, 8:58 p.m.