niaaml-optimize - Man Page

niaaml-optimize – optimizes a NiaAML pipeline on a given dataset

Synopsis

niaaml optimize [Options] DATA_CSV_FILE

--has-header, --no-has-header: [default: has-header]
--ignore-columns: INTEGER
--classifiers: TEXT [default: AdaBoost, Bagging, MultiLayerPerceptron, RandomForest, ExtremelyRandomizedTrees, LinearSVC]
--feature-selection-algorithms: TEXT [default: SelectKBest, SelectPercentile, ParticleSwarmOptimization, VarianceThreshold]
--feature-transform-algorithms: TEXT [default: Normalizer, StandardScaler]
--categorical-features-encoder: TEXT [default: OneHotEncoder]
--imputer: TEXT [default: None]
--fitness-name: TEXT [default: Accuracy]
--pipeline-population-size: INTEGER [default: 15]
--inner-population-size: INTEGER [default: 15]
--number-of-pipeline-evaluations: INTEGER [default: 100]
--number-of-inner-evaluations: INTEGER [default: 100]
--optimization-algorithm: TEXT [default: ParticleSwarmAlgorithm]
--inner-optimization-algorithm: TEXT [default: None]
--result-file: TEXT [default: pipeline.ppln]
-h, --help: Show a help message and exit

September 2024