Mercurial > repos > jay > ml_tool
view ml_tool/ml_tool.xml @ 0:e94586e24004 draft default tip
planemo upload for repository https://github.com/jaidevjoshi83/MicroBiomML commit 5ef78d4decc95ac107c468499328e7f086289ff9-dirty
| author | jay |
|---|---|
| date | Tue, 17 Feb 2026 10:52:24 +0000 |
| parents | |
| children |
line wrap: on
line source
<tool id="ml_tool" name="ML Tool" version="0.1.0" python_template_version="3.10"> <description>Create machine learning classification models using multiple algorithms for microbiome data analysis.</description> <requirements> <requirement type="package" version="3.10">Python</requirement> <requirement type="package" version="3.3">pycaret</requirement> <requirement type="package" version="2.1.1">xgboost</requirement> <requirement type="package" version="0.1.18">hdlib</requirement> </requirements> <command detect_errors="exit_code"><![CDATA[ python '$__tool_directory__/ml_tool.py' --algo $SelMLAlgo.MLAlgo --data_file $input1 --metadata_file $input2 --output_tabular $output1 --output_html $output2 --target_label $column_label #if $setup.advanced_setup == 'settings' --setup '{"session_id":$setup.session_id, "fold":$setup.fold, "log_experiment":$setup.log_experiment, "train_size":$setup.train_size, "data_split_shuffle":$setup.data_split_shuffle, "data_split_stratify":$setup.data_split_stratify, "normalize":$setup.normalize, "transformation":$setup.transformation, "remove_outliers":$setup.remove_outliers, "outliers_threshold":$setup.outliers_threshold, "pca":$setup.pca}' #end if #if $setup.advanced_setup == 'default' --setup '{}' #end if #if $drop_columns.advanced_setup == 'settings' --dp_columns '$drop_columns.columns_to_drop' #end if #if $SelMLAlgo.MLAlgo == 'rbfsvm' #if $SelMLAlgo.settings.advanced == "settings" #if $SelMLAlgo.settings.settings.advanced == "custom" --custom_para '{"C": $SelMLAlgo.settings.settings.C, "break_ties":$SelMLAlgo.settings.settings.break_ties, "cache_size":$SelMLAlgo.settings.settings.cache_size, "class_weight":$SelMLAlgo.settings.settings.class_weight, "coef0":$SelMLAlgo.settings.settings.coef0, "decision_function_shape":"$SelMLAlgo.settings.settings.decision_function_shape", "degree":$SelMLAlgo.settings.settings.degree, "gamma":"$SelMLAlgo.settings.settings.gamma", "kernel":"$SelMLAlgo.settings.settings.kernel", "max_iter":$SelMLAlgo.settings.settings.max_iter, "probability":$SelMLAlgo.settings.settings.probability, "random_state":$SelMLAlgo.settings.settings.random_state, "shrinking":$SelMLAlgo.settings.settings.shrinking,"tol":$SelMLAlgo.settings.settings.tol, "verbose":$SelMLAlgo.settings.settings.verbose}' #end if #if $SelMLAlgo.settings.settings.advanced == "tune" --param_file"$tune_param_file" #end if #end if #end if #if $SelMLAlgo.MLAlgo == 'gbc' #if $SelMLAlgo.settings.advanced == "settings" #if $SelMLAlgo.settings.settings.advanced == "custom" --custom_para '{"ccp_alpha":$SelMLAlgo.settings.ccp_alpha, "criterion":"$SelMLAlgo.settings.criterion", "init":$SelMLAlgo.settings.init, "learning_rate":$SelMLAlgo.settings.learning_rate, "loss":"$SelMLAlgo.settings.loss", "max_depth":$SelMLAlgo.settings.max_depth, "max_features":$SelMLAlgo.settings.max_features, "max_leaf_nodes":$SelMLAlgo.settings.max_leaf_nodes, "min_impurity_decrease":$SelMLAlgo.settings.min_impurity_decrease, "min_samples_leaf":$SelMLAlgo.settings.min_samples_leaf, "min_samples_split":$SelMLAlgo.settings.min_samples_split, "min_weight_fraction_leaf":$SelMLAlgo.settings.min_weight_fraction_leaf, "n_estimators":$SelMLAlgo.settings.n_estimators, "n_iter_no_change":$SelMLAlgo.settings.n_iter_no_change, "random_state":$SelMLAlgo.settings.random_state, "subsample":$SelMLAlgo.settings.subsample, "tol":$SelMLAlgo.settings.tol, "validation_fraction":$SelMLAlgo.settings.validation_fraction, "verbose":$SelMLAlgo.settings.verbose, "warm_start":$SelMLAlgo.settings.warm_start}' #end if #if $SelMLAlgo.settings.settings.advanced == "tune" --param_file "$tune_param_file" #end if #end if #end if #if $SelMLAlgo.MLAlgo == 'dt' #if $SelMLAlgo.settings.advanced == "settings" #if $SelMLAlgo.settings.settings.advanced == "custom" --custom_para '{"ccp_alpha": $SelMLAlgo.settings.settings.ccp_alpha, "class_weight": $SelMLAlgo.settings.settings.class_weight, "criterion":"$SelMLAlgo.settings.settings.criterion", "max_depth":$SelMLAlgo.settings.settings.max_depth, "max_features":$SelMLAlgo.settings.settings.max_features, "max_leaf_nodes":$SelMLAlgo.settings.settings.max_leaf_nodes, "min_impurity_decrease":$SelMLAlgo.settings.settings.min_impurity_decrease, "min_samples_leaf":$SelMLAlgo.settings.settings.min_samples_leaf, "min_samples_split":$SelMLAlgo.settings.settings.min_samples_split, "min_weight_fraction_leaf":$SelMLAlgo.settings.settings.min_weight_fraction_leaf, "monotonic_cst":$SelMLAlgo.settings.settings.monotonic_cst, "random_state":$SelMLAlgo.settings.settings.random_state, "splitter":"$SelMLAlgo.settings.settings.splitter"}' #end if #if $SelMLAlgo.settings.settings.advanced == "tune" --param_file "$tune_param_file" #end if #end if #end if #if $SelMLAlgo.MLAlgo == 'rf' #if $SelMLAlgo.settings.advanced == "settings" #if $SelMLAlgo.settings.settings.advanced == "custom" --custom_para '{"bootstrap": $SelMLAlgo.settings.settings.bootstrap, "ccp_alpha": $SelMLAlgo.settings.settings.ccp_alpha, "class_weight": $SelMLAlgo.settings.settings.class_weight, "criterion": "$SelMLAlgo.settings.settings.criterion", "max_depth": $SelMLAlgo.settings.settings.max_depth, "max_features": "$SelMLAlgo.settings.settings.max_features", "max_leaf_nodes": $SelMLAlgo.settings.settings.max_leaf_nodes, "max_samples": $SelMLAlgo.settings.settings.max_samples, "min_impurity_decrease": $SelMLAlgo.settings.settings.min_impurity_decrease, "min_samples_leaf": $SelMLAlgo.settings.settings.min_samples_leaf, "min_samples_split": $SelMLAlgo.settings.settings.min_samples_split, "min_weight_fraction_leaf": $SelMLAlgo.settings.settings.min_weight_fraction_leaf, "monotonic_cst": $SelMLAlgo.settings.settings.monotonic_cst, "n_estimators": $SelMLAlgo.settings.settings.n_estimators, "oob_score": $SelMLAlgo.settings.settings.oob_score, "random_state": $SelMLAlgo.settings.settings.random_state, "verbose": $SelMLAlgo.settings.settings.verbose, "warm_start": $SelMLAlgo.settings.settings.warm_start }' #end if #if $SelMLAlgo.settings.settings.advanced == "tune" --param_file "$tune_param_file" #end if #end if #end if #if $SelMLAlgo.MLAlgo == 'lr' #if $SelMLAlgo.settings.advanced == "settings" #if $SelMLAlgo.settings.settings.advanced == "custom" --custom_para '{"C":$SelMLAlgo.settings.settings.C, "dual":$SelMLAlgo.settings.settings.dual, "fit_intercept":$SelMLAlgo.settings.settings.fit_intercept, "intercept_scaling":$SelMLAlgo.settings.settings.intercept_scaling, "l1_ratio":$SelMLAlgo.settings.settings.l1_ratio, "max_iter":$SelMLAlgo.settings.settings.max_iter, "multi_class":"$SelMLAlgo.settings.settings.multi_class", "penalty":"$SelMLAlgo.settings.settings.penalty", "random_state":$SelMLAlgo.settings.settings.random_state, "solver":"$SelMLAlgo.settings.settings.solver", "tol":$SelMLAlgo.settings.settings.tol, "verbose":$SelMLAlgo.settings.settings.verbose, "warm_start":$SelMLAlgo.settings.settings.warm_start}' #end if #if $SelMLAlgo.settings.settings.advanced == "tune" --param_file "$tune_param_file" #end if #end if #end if #if $SelMLAlgo.MLAlgo == 'knn' #if $SelMLAlgo.settings.advanced == "settings" #if $SelMLAlgo.settings.settings.advanced == "custom" --custom_para '{"algorithm": "$SelMLAlgo.settings.settings.algorithm", "leaf_size":$SelMLAlgo.settings.settings.leaf_size, "metric": "$SelMLAlgo.settings.settings.metric", "n_neighbors": $SelMLAlgo.settings.settings.n_neighbors, "p":$SelMLAlgo.settings.settings.p, "weights":"$SelMLAlgo.settings.settings.weights"}' #end if #if $SelMLAlgo.settings.settings.advanced == "tune" --param_file "$tune_param_file" #end if #end if #end if #if $SelMLAlgo.MLAlgo == 'nb' #if $SelMLAlgo.settings.advanced == "settings" #if $SelMLAlgo.settings.settings.advanced == "custom" --custom_para '{"priors": $SelMLAlgo.settings.settings.priors, "var_smoothing":$SelMLAlgo.settings.settings.var_smoothing}' #end if #if $SelMLAlgo.settings.settings.advanced == "tune" --param_file "$tune_param_file" #end if #end if #end if #if $SelMLAlgo.MLAlgo == 'xgboost' #if $SelMLAlgo.settings.advanced == "settings" #if $SelMLAlgo.settings.settings.advanced == "custom" --custom_para '{"objective": "$SelMLAlgo.settings.settings.objective", "base_score": $SelMLAlgo.settings.settings.base_score, "booster": "$SelMLAlgo.settings.settings.booster", "colsample_bylevel": $SelMLAlgo.settings.settings.colsample_bylevel, "colsample_bynode": $SelMLAlgo.settings.settings.colsample_bynode, "colsample_bytree": $SelMLAlgo.settings.settings.colsample_bytree, "early_stopping_rounds": $SelMLAlgo.settings.settings.early_stopping_rounds, "enable_categorical": $SelMLAlgo.settings.settings.enable_categorical, "eval_metric": $SelMLAlgo.settings.settings.eval_metric, "feature_types": $SelMLAlgo.settings.settings.feature_types, "gamma": $SelMLAlgo.settings.settings.gamma, "grow_policy": $SelMLAlgo.settings.settings.grow_policy, "importance_type":$SelMLAlgo.settings.settings.importance_type, "interaction_constraints":$SelMLAlgo.settings.settings.interaction_constraints, "learning_rate":$SelMLAlgo.settings.settings.learning_rate, "max_bin":$SelMLAlgo.settings.settings.max_bin, "max_cat_threshold":$SelMLAlgo.settings.settings.max_cat_threshold, "max_cat_to_onehot":$SelMLAlgo.settings.settings.max_cat_to_onehot, "max_delta_step":$SelMLAlgo.settings.settings.max_delta_step, "max_depth":$SelMLAlgo.settings.settings.max_depth, "max_leaves":$SelMLAlgo.settings.settings.max_leaves, "min_child_weight":$SelMLAlgo.settings.settings.min_child_weight, "monotone_constraints":$SelMLAlgo.settings.settings.monotone_constraints, "multi_strategy":$SelMLAlgo.settings.settings.multi_strategy, "n_estimators":$SelMLAlgo.settings.settings.n_estimators, "num_parallel_tree":$SelMLAlgo.settings.settings.num_parallel_tree, "random_state":$SelMLAlgo.settings.settings.random_state, "reg_alpha":$SelMLAlgo.settings.settings.reg_alpha, "reg_lambda":$SelMLAlgo.settings.settings.reg_lambda, "sampling_method":"$SelMLAlgo.settings.settings.sampling_method", "scale_pos_weight":$SelMLAlgo.settings.settings.scale_pos_weight, "subsample":$SelMLAlgo.settings.settings.subsample, "tree_method":"$SelMLAlgo.settings.settings.tree_method", "validate_parameters":$SelMLAlgo.settings.settings.validate_parameters, "verbosity":$SelMLAlgo.settings.settings.verbosity}' #end if #if $SelMLAlgo.settings.settings.advanced == "tune" --param_file "$tune_param_file" #end if #end if #end if #if $SelMLAlgo.MLAlgo == 'lightgbm' #if $SelMLAlgo.settings.advanced == "settings" #if $SelMLAlgo.settings.settings.advanced == "custom" --custom_para '{"boosting_type":"$SelMLAlgo.settings.settings.boosting_type", "class_weight":$SelMLAlgo.settings.settings.class_weight, "colsample_bytree":$SelMLAlgo.settings.settings.colsample_bytree, "importance_type":"$SelMLAlgo.settings.settings.importance_type", "learning_rate":$SelMLAlgo.settings.settings.learning_rate, "max_depth":$SelMLAlgo.settings.settings.max_depth, "min_child_samples":$SelMLAlgo.settings.settings.min_child_samples, "min_child_weight":$SelMLAlgo.settings.settings.min_child_weight, "min_split_gain":$SelMLAlgo.settings.settings.min_split_gain, "n_estimators":$SelMLAlgo.settings.settings.n_estimators, "num_leaves":$SelMLAlgo.settings.settings.num_leaves, "objective":"$SelMLAlgo.settings.settings.objective", "random_state":$SelMLAlgo.settings.settings.random_state, "reg_alpha":$SelMLAlgo.settings.settings.reg_alpha, "reg_lambda":$SelMLAlgo.settings.settings.reg_lambda, "subsample":$SelMLAlgo.settings.settings.subsample, "subsample_for_bin":$SelMLAlgo.settings.settings.subsample_for_bin, "subsample_freq":$SelMLAlgo.settings.settings.subsample_freq}' #end if #if $SelMLAlgo.settings.settings.advanced == "tune" --param_file "$tune_param_file" #end if #end if #end if #if $SelMLAlgo.MLAlgo == 'svm' #if $SelMLAlgo.settings.advanced == "settings" #if $SelMLAlgo.settings.settings.advanced == "custom" --custom_para '{"alpha":$SelMLAlgo.settings.settings.alpha, "average":$SelMLAlgo.settings.settings.average, "early_stopping":$SelMLAlgo.settings.settings.early_stopping, "epsilon":$SelMLAlgo.settings.settings.epsilon, "eta0":$SelMLAlgo.settings.settings.eta0, "fit_intercept":$SelMLAlgo.settings.settings.fit_intercept, "l1_ratio":$SelMLAlgo.settings.settings.l1_ratio, "learning_rate":"$SelMLAlgo.settings.settings.learning_rate", "loss":"$SelMLAlgo.settings.settings.loss", "max_iter":$SelMLAlgo.settings.settings.max_iter, "n_iter_no_change":$SelMLAlgo.settings.settings.n_iter_no_change, "penalty":"$SelMLAlgo.settings.settings.penalty", "power_t":$SelMLAlgo.settings.settings.power_t, "random_state":$SelMLAlgo.settings.settings.random_state, "shuffle":$SelMLAlgo.settings.settings.shuffle, "tol":$SelMLAlgo.settings.settings.tol, "validation_fraction":$SelMLAlgo.settings.settings.validation_fraction, "verbose":$SelMLAlgo.settings.settings.verbose, "warm_start":$SelMLAlgo.settings.settings.warm_start}' #end if #if $SelMLAlgo.settings.settings.advanced == "tune" --param_file "$tune_param_file" #end if #end if #end if #if $SelMLAlgo.MLAlgo == 'hdc' #if $SelMLAlgo.settings.advanced == "settings" #if $SelMLAlgo.settings.settings.advanced == "custom" --custom_para '{"dimensionality":$SelMLAlgo.settings.settings.dimensionality, "levels":$SelMLAlgo.settings.settings.levels, "retrain":$SelMLAlgo.settings.settings.retrain}' #end if #if $SelMLAlgo.settings.settings.advanced == "tune" --param_file "$tune_param_file" #end if #end if #end if ]]></command> <inputs> <param name="input1" type="data" format="tabular" label="Training Data" optional="false" argument= "--data_file"/> <param name="input2" type="data" format="tabular" label="Metadata" optional="false" argument= "--metadata_file"/> <param name="column_label" type='data_column' data_ref="input2" label="Target Column (Class Label)" multiple="False" use_header_names="true" help="Select the target column for classification (e.g., phenotype, class, label)."></param> <!-- Index column options removed --> <conditional name='drop_columns' > <param name="advanced_setup" type="select" label="Drop Columns from Training Data"> <option value="default" selected="true">Do Not Drop Columns</option> <option value="settings">Drop Columns</option> </param> <when value="default"> </when> <when value="settings"> <param name="columns_to_drop" type='data_column' data_ref="input1" label="Columns to Drop from Training Data" argument="--dp_columns" multiple="true" use_header_names="true" help="Select the columns to drop from the training data."> </param> </when> </conditional> <conditional name='setup' > <param name="advanced_setup" type="select" label="Advanced Setup (Advanced setup is not required for HDC classifier)"> <option value="default" selected="true">Use Default Setup</option> <option value="settings">Use Advanced Setup</option> </param> <when value="default"> </when> <when value="settings"> <!-- Target Column selector removed as requested --> <!-- Session ID (Optional) --> <param name="session_id" type="integer" value="123" label="Session ID (for Reproducibility)" help="The session ID for reproducibility." optional="true" /> <!-- Feature Selection (Optional) --> <!-- Number of Folds (Optional) --> <param name="fold" type="integer" value="10" label="Number of Folds for Cross-Validation" help="The number of folds for cross-validation." optional="true" /> <!-- Log Experiment (Optional) --> <param name="log_experiment" type="boolean" value="false" label="Log Experiment" help="Whether to log the experiment." optional="true" /> <!-- Training Set Size (Optional) --> <param name="train_size" type="float" value="0.7" label="Training Set Size" help="The proportion of the dataset to include in the training split." optional="true" /> <!-- Shuffle Data (Optional) --> <param name="data_split_shuffle" type="boolean" value="True" label="Shuffle Data" help="Whether to shuffle data before splitting." optional="true" /> <!-- Stratify Data (Optional) --> <param name="data_split_stratify" type="boolean" value="False" label="Stratify Data" help="Whether to stratify data during splitting." optional="true" /> <!-- Normalize Data (Optional) --> <param name="normalize" type="boolean" value="false" label="Normalize Data" help="Whether to normalize the data." optional="true" /> <!-- Power Transformation (Optional) --> <param name="transformation" type="boolean" value="false" label="Apply Power Transformation" help="Whether to apply power transformation." optional="true" /> <!-- Remove Outliers (Optional) --> <param name="remove_outliers" type="boolean" value="false" label="Remove Outliers" help="Whether to remove outliers from the data." optional="true" /> <!-- Outlier Threshold (Optional) --> <param name="outliers_threshold" type="float" value="0.05" label="Outlier Threshold" help="The threshold for removing outliers." optional="true" /> <!-- PCA --> <param name="pca" type="boolean" label="Apply PCA" help="Whether to apply PCA for dimensionality reduction."/> </when> </conditional> <conditional name='SelMLAlgo' > <param name="MLAlgo" type="select" label="Machine Learning Algorithm" argument=""> <option value="lr">Logistic Regression Classifier</option> <option value="knn">K-Nearest Neighbors Classifier</option> <option value="nb">Gaussian Naive Bayes Classifier</option> <option value="dt">Decision Tree Classifier</option> <option value="svm">Support Vector Machine Classifier</option> <option value="rbfsvm">RBF Support Vector Machine Classifier</option> <option value="rf">Random Forest Classifier</option> <option value="gbc">Gradient Boosting Classifier</option> <option value="xgboost">XGBoost Classifier</option> <option value="lightgbm">LightGBM Classifier</option> <option value="hdc">Vector Symbolic Architecture-based Classifier</option> </param> <when value="lr"> <conditional name="settings"> <param name="advanced" type="select" label="Use Advanced Settings to Create Model."> <option value="default" selected="true">Use Default Setup</option> <option value="settings">Use Advanced Setup</option> </param> <when value="default"> </when> <when value="settings"> <conditional name="settings"> <param name="advanced" type="select" label="Specify Advanced Parameters."> <option value="custom" selected="true">Create Model with Custom Parameters</option> <option value="tune">Tune Model</option> </param> <when value="custom"> <param name="class_weight" type="text" value="null" label="Class Weight" optional="true"/> <param name="dual" type="boolean" value="false" label="Dual Formulation"/> <param name="fit_intercept" type="boolean" value="True" label="Fit Intercept"/> <param name="intercept_scaling" type="float" value="1" label="Intercept Scaling"/> <param name="l1_ratio" type="text" value="null" label="L1 Ratio" optional="true"/> <param name="max_iter" type="integer" value="1000" label="Max Iterations"/> <param name="multi_class" type="select" label="Multi-Class"> <option value="auto">Auto</option> <option value="ovr">One-vs-Rest</option> <option value="multinomial">Multinomial</option> </param> <param name="n_jobs" type="text" value="null" label="Number of Jobs" optional="true"/> <param name="penalty" type="select" label="Penalty"> <option value="l2">L2</option> <option value="l1">L1</option> <option value="elasticnet">Elastic Net</option> <option value="null">None</option> </param> <param name="random_state" type="integer" value="4131" label="Random State"/> <param name="solver" type="select" label="Solver"> <option value="lbfgs">LBFGS</option> <option value="liblinear">Liblinear</option> <option value="newton-cg">Newton-CG</option> <option value="sag">SAG</option> <option value="saga">SAGA</option> </param> <param name="tol" type="float" value="0.0001" label="Tolerance"/> <param name="verbose" type="boolean" value="false" label="Verbose"/> <param name="warm_start" type="boolean" value="false" label="Warm Start"/> </when> <when value="tune"> <param name="tune_param_file" type="data" format="txt" label="Training data" optional="false" argument= "--input_data"/> </when> </conditional> </when> </conditional> </when> <when value="knn"> <conditional name="settings"> <param name="advanced" type="select" label="Use Advanced Settings to Create Model."> <option value="default" selected="true">Use Default Setup</option> <option value="settings">Use Advanced Setup</option> </param> <when value="default"> </when> <when value="settings"> <conditional name="settings"> <param name="advanced" type="select" label="Specify Advanced Parameters."> <option value="custom" selected="true">Create Model with Custom Parameters</option> <option value="tune">Tune Model</option> </param> <when value="custom"> <param name="algorithm" type="select" label="Algorithm"> <option value="auto" selected='true' >Auto</option> <option value="ball_tree">Ball Tree</option> <option value="kd_tree">KD Tree</option> <option value="brute">Brute</option> </param> <param name="leaf_size" type="integer" value="30" label="Leaf Size"/> <param name="metric" type="select" label="Metric"> <option value="minkowski" selected='true' >Minkowski</option> <option value="euclidean">Euclidean</option> <option value="manhattan">Manhattan</option> <option value="chebyshev">Chebyshev</option> <option value="cosine">Cosine</option> </param> <param name="n_neighbors" type="integer" value="5" label="Number of Neighbors"/> <param name="p" type="integer" value="2" label="Power Parameter for Minkowski"/> <param name="weights" type="select" label="Weight Function"> <option value="uniform" selected='true'>Uniform</option> <option value="distance">Distance</option> </param> </when> <when value="tune"> <param name="tune_param_file" type="data" format="txt" label="Training data" optional="false" argument= "--input_data"/> </when> </conditional> </when> </conditional> </when> <when value="nb"> <conditional name="settings"> <param name="advanced" type="select" label="Use Advanced Settings to Create Model."> <option value="default" selected="true">Use Default Setup</option> <option value="settings">Use Advanced Setup</option> </param> <when value="default"> </when> <when value="settings"> <conditional name="settings"> <param name="advanced" type="select" label="Specify Advanced Parameters."> <option value="custom" selected="true">Create Model with Custom Parameters</option> <option value="tune">Tune Model</option> </param> <when value="custom"> <param name="priors" type="text" value="null" label="Class Priors" optional="true"/> <param name="var_smoothing" type="float" value="1e-09" label="Variance Smoothing"/> </when> <when value="tune"> <param name="tune_param_file" type="data" format="txt" label="Training data" optional="false" argument= "--input_data"/> </when> </conditional> </when> </conditional> </when> <when value="dt"> <conditional name="settings"> <param name="advanced" type="select" label="Use Advanced Settings to Create Model."> <option value="default" selected="true">Use Default Setup</option> <option value="settings">Use Advanced Setup</option> </param> <when value="default"> </when> <when value="settings"> <conditional name="settings"> <param name="advanced" type="select" label="Specify Advanced Parameters."> <option value="custom" selected="true">Create Model with Custom Parameters</option> <option value="tune">Tune Model</option> </param> <when value="custom"> <param name="ccp_alpha" type="float" value="0.0" label="Complexity Parameter Alpha"/> <param name="class_weight" type="text" value="null" label="Class Weight" optional="true"/> <param name="criterion" type="select" label="Criterion"> <option value="gini">Gini</option> <option value="entropy">Entropy</option> </param> <param name="max_depth" type="text" value="null" label="Max Depth" optional="true"/> <param name="max_features" type="text" value="null" label="Max Features" optional="true"/> <param name="max_leaf_nodes" type="text" value="null" label="Max Leaf Nodes" optional="true"/> <param name="min_impurity_decrease" type="float" value="0.0" label="Min Impurity Decrease"/> <param name="min_samples_leaf" type="integer" value="1" label="Min Samples Leaf"/> <param name="min_samples_split" type="integer" value="2" label="Min Samples Split"/> <param name="min_weight_fraction_leaf" type="float" value="0.0" label="Min Weight Fraction Leaf"/> <param name="monotonic_cst" type="text" value="null" label="Monotonic Constraints" optional="true"/> <param name="random_state" type="integer" value="4131" label="Random State"/> <param name="splitter" type="select" label="Splitter"> <option value="best">Best</option> <option value="random">Random</option> </param> </when> <when value="tune"> <param name="tune_param_file" type="data" format="txt" label="Training data" optional="false" argument= "--input_data"/> </when> </conditional> </when> </conditional> </when> <when value="svm"> <conditional name="settings"> <param name="advanced" type="select" label="Use Advanced Settings to Create Model."> <option value="default" selected="true">Use Default Setup</option> <option value="settings">Use Advanced Setup</option> </param> <when value="default"> </when> <when value="settings"> <conditional name="settings"> <param name="advanced" type="select" label="Specify Advanced Parameters."> <option value="custom" selected="true">Create Model with Custom Parameters</option> <option value="tune">Tune Model</option> </param> <when value="custom"> <param name="alpha" type="float" value="0.0001" label="Alpha" /> <param name="average" type="boolean" value="False" label="Average" /> <param name="early_stopping" type="boolean" value="False" label="Early Stopping" /> <param name="epsilon" type="float" value="0.1" label="Epsilon" /> <param name="eta0" type="float" value="0.001" label="Eta0" /> <param name="fit_intercept" type="boolean" value="True" label="Fit Intercept" /> <param name="l1_ratio" type="float" value="0.15" label="L1 Ratio" /> <param name="learning_rate" type="select" multiple="true" label="Learning Rate"> <option value="constant">Constant</option> <option value="optimal" selected="true" >Optimal</option> <option value="invscaling">Invscaling</option> </param> <param name="loss" type="select" multiple="true" label="Loss"> <option selected="true" value="hinge">Hinge</option> <option value="log">Log</option> <option value="modified_huber">Modified Huber</option> <option value="squared_hinge">Squared Hinge</option> <option value="perceptron">Perceptron</option> </param> <param name="max_iter" type="integer" value="1000" label="Max Iterations" /> <param name="n_iter_no_change" type="integer" value="5" label="Number of Iterations without Change" /> <param name="n_jobs" type="integer" value="-1" label="Number of Jobs" /> <param name="penalty" type="select" multiple="true" label="Penalty"> <option selected="true" value="l2">L2</option> <option value="l1">L1</option> <option value="elasticnet">Elastic Net</option> </param> <param name="power_t" type="float" value="0.5" label="Power T" /> <param name="random_state" type="integer" value="4286" label="Random State" /> <param name="shuffle" type="boolean" value="True" label="Shuffle" /> <param name="tol" type="float" value="0.001" label="Tolerance" /> <param name="validation_fraction" type="float" value="0.1" label="Validation Fraction" /> <param name="verbose" type="boolean" value="False" label="Verbose" /> <param name="warm_start" type="boolean" value="False" label="Warm Start" /> </when> <when value="tune"> <param name="tune_param_file" type="data" format="txt" label="Training data" optional="false" argument= "--input_data"/> </when> </conditional> </when> </conditional> </when> <when value="rbfsvm"> <conditional name="settings"> <param name="advanced" type="select" label="Use Advanced Settings to Create Model."> <option value="default" selected="true">Use Default Setup</option> <option value="settings">Use Advanced Setup</option> </param> <when value="default"> </when> <when value="settings"> <conditional name="settings"> <param name="advanced" type="select" label="Specify Advanced Parameters."> <option value="custom" selected="true">Create Model with Custom Parameters</option> <option value="tune">Tune Model</option> </param> <when value="custom"> <param name="C" type="float" value="1.0" /> <param name="break_ties" type="boolean" value="False" /> <param name="cache_size" type="float" value="200" /> <param name="class_weight" type="text" value="null" /> <param name="coef0" type="float" value="0.0" /> <param name="decision_function_shape" type="select" value="ovr"> <option value="ovr">ovr</option> <option value="ovo">ovo</option> </param> <param name="degree" type="integer" value="3" /> <param name="gamma" type="select" value="auto"> <option value="auto">auto</option> <option value="scale">scale</option> </param> <param name="kernel" type="select" value="rbf"> <option value="linear">linear</option> <option value="poly">poly</option> <option value="rbf">rbf</option> <option value="sigmoid">sigmoid</option> </param> <param name="max_iter" type="integer" value="-1" /> <param name="probability" type="boolean" value="True" /> <param name="random_state" type="integer" value="4131" /> <param name="shrinking" type="boolean" value="True" /> <param name="tol" type="float" value="0.001" /> <param name="verbose" type="boolean" value="False" /> </when> <when value="tune"> <param name="tune_param_file" type="data" format="txt" label="Training data" optional="false" argument= "--input_data"/> </when> </conditional> </when> </conditional> </when> <when value="rf"> <conditional name="settings"> <param name="advanced" type="select" label="Use Advanced Settings to Create Model."> <option value="default" selected="true">Use Default Setup</option> <option value="settings">Use Advanced Setup</option> </param> <when value="default"> </when> <when value="settings"> <conditional name="settings"> <param name="advanced" type="select" label="Specify Advanced Parameters."> <option value="custom" selected="true">Create Model with Custom Parameters</option> <option value="tune">Tune Model</option> </param> <when value="custom"> <param name="bootstrap" type="boolean" value="True" label="Bootstrap Samples"/> <param name="ccp_alpha" type="float" value="0.0" label="Complexity Parameter Alpha"/> <param name="class_weight" type="text" value="null" label="Class Weight" optional="true"/> <param name="criterion" type="select" label="Criterion"> <option value="gini">Gini</option> <option value="entropy">Entropy</option> </param> <param name="max_depth" type="text" value="null" label="Max Depth" optional="true"/> <param name="max_features" type="select" label="Max Features"> <option value="sqrt">Sqrt</option> <option value="log2">Log2</option> <option value="null">None</option> </param> <param name="max_leaf_nodes" type="text" value="null" label="Max Leaf Nodes" optional="true"/> <param name="max_samples" type="text" value="null" label="Max Samples" optional="true"/> <param name="min_impurity_decrease" type="float" value="0.0" label="Min Impurity Decrease"/> <param name="min_samples_leaf" type="integer" value="1" label="Min Samples Leaf"/> <param name="min_samples_split" type="integer" value="2" label="Min Samples Split"/> <param name="min_weight_fraction_leaf" type="float" value="0.0" label="Min Weight Fraction Leaf"/> <param name="monotonic_cst" type="text" value="null" label="Monotonic Constraints" optional="true"/> <param name="n_estimators" type="integer" value="100" label="Number of Estimators"/> <param name="oob_score" type="boolean" value="False" label="Out-of-Bag Score"/> <param name="random_state" type="integer" value="4131" label="Random State"/> <param name="verbose" type="boolean" value="True" label="Verbose"/> <param name="warm_start" type="boolean" value="False" label="Warm Start"/> </when> <when value="tune"> <param name="tune_param_file" type="data" format="txt" label="Training data" optional="false" argument= "--input_data"/> </when> </conditional> </when> </conditional> </when> <when value="gbc"> <conditional name="settings"> <param name="advanced" type="select" label="Use Advanced Settings to Create Model."> <option value="default" selected="true">Use Default Setup</option> <option value="settings">Use Advanced Setup</option> </param> <when value="default"> </when> <when value="settings"> <conditional name="settings"> <param name="advanced" type="select" label="Specify Advanced Parameters."> <option value="custom" selected="true">Create Model with Custom Parameters</option> <option value="tune">Tune Model</option> </param> <when value="custom"> <param name="ccp_alpha" type="float" value="0.0" label="Complexity Parameter (ccp_alpha)"/> <param name="criterion" type="select" label="Criterion"> <option value="friedman_mse">Friedman MSE</option> <option value="squared_error">Squared Error</option> <option value="mae">Mean Absolute Error</option> </param> <param name="init" type="text" value="null" label="Init" optional="true"/> <param name="learning_rate" type="float" value="0.1" label="Learning Rate"/> <param name="loss" type="select" label="Loss Function"> <option value="log_loss">Log Loss</option> <option value="deviance">Deviance</option> <option value="exponential">Exponential</option> </param> <param name="max_depth" type="integer" value="3" label="Max Depth"/> <param name="max_features" type="text" value="null" label="Max Features" optional="true"/> <param name="max_leaf_nodes" type="text" value="null" label="Max Leaf Nodes" optional="true"/> <param name="min_impurity_decrease" type="float" value="0.0" label="Min Impurity Decrease"/> <param name="min_samples_leaf" type="integer" value="1" label="Min Samples Leaf"/> <param name="min_samples_split" type="integer" value="2" label="Min Samples Split"/> <param name="min_weight_fraction_leaf" type="float" value="0.0" label="Min Weight Fraction Leaf"/> <param name="n_estimators" type="integer" value="100" label="Number of Estimators"/> <param name="n_iter_no_change" type="text" value="null" label="Number of Iterations with No Change" optional="true"/> <param name="random_state" type="integer" value="4131" label="Random State"/> <param name="subsample" type="float" value="1.0" label="Subsample"/> <param name="tol" type="float" value="0.0001" label="Tolerance"/> <param name="validation_fraction" type="float" value="0.1" label="Validation Fraction"/> <param name="verbose" type="boolean" value="True" label="Verbose"/> <param name="warm_start" type="boolean" value="False" label="Warm Start"/> </when> <when value="tune"> <param name="tune_param_file" type="data" format="txt" label="Training data" optional="false" argument= "--input_data"/> </when> </conditional> </when> </conditional> </when> <when value="xgboost"> <conditional name="settings"> <param name="advanced" type="select" label="Use Advanced Settings to Create Model."> <option value="default" selected="true">Use Default Setup</option> <option value="settings">Use Advanced Setup</option> </param> <when value="default"> </when> <when value="settings"> <conditional name="settings"> <param name="advanced" type="select" label="Specify Advanced Parameters."> <option value="custom" selected="true">Create Model with Custom Parameters</option> <option value="tune">Tune Model</option> </param> <when value="custom"> <param name="objective" type="select" value="reg:squarederror" label="Objective Function"> <option value="reg:squarederror">Regression with squared loss</option> <option value="reg:squaredlogerror">Regression with squared log loss</option> <option value="reg:logistic">Logistic regression, output probability</option> <option value="reg:pseudohubererror">Regression with Pseudo Huber loss</option> <option value="reg:absoluteerror">Regression with L1 error</option> <option value="reg:quantileerror">Quantile loss (pinball loss)</option> <option value="binary:logistic">Logistic regression for binary classification (probability)</option> <option value="binary:logitraw">Logistic regression for binary classification (score before transformation)</option> <option value="binary:hinge">Hinge loss for binary classification (0 or 1 predictions)</option> <option value="count:poisson">Poisson regression for count data</option> <option value="survival:cox">Cox regression for right censored survival time data</option> <option value="survival:aft">Accelerated failure time model for censored survival time data</option> <option value="multi:softmax">Multiclass classification using softmax</option> <option value="multi:softprob">Multiclass classification with probability output</option> <option value="rank:ndcg">Pair-wise ranking using LambdaMART to maximize NDCG</option> <option value="rank:map">Pair-wise ranking using LambdaMART to maximize MAP</option> <option value="rank:pairwise">Pair-wise ranking using LambdaRank</option> <option value="reg:gamma">Gamma regression with log-link</option> <option value="reg:tweedie">Tweedie regression with log-link</option> </param> <param name="base_score" type="float" label="Base Score" value="0.5" optional="true" /> <param name="colsample_bylevel" type="float" label="Colsample by Level" value="1" /> <param name="colsample_bynode" type="float" label="Colsample by Node" value="1" optional="true" /> <param name="colsample_bytree" type="float" label="Colsample by Tree" value="1" optional="true" /> <param name="early_stopping_rounds" type="integer" label="Early Stopping Rounds" value="0" optional="true" /> <param name="enable_categorical" type="boolean" label="Enable Categorical" value="False" /> <param name="eval_metric" type="text" label="Eval Metric" value="null" optional="true" /> <param name="feature_types" type="text" label="Feature Types" value="null" optional="true" /> <param name="gamma" type="float" label="Gamma" value="0" optional="true" /> <param name="grow_policy" type="text" label="Grow Policy" value="null" optional="true" /> <param name="importance_type" type="text" label="Importance Type" value="null" optional="true" /> <param name="interaction_constraints" type="text" label="Interaction Constraints" value="null" optional="true" /> <param name="learning_rate" type="float" label="Learning Rate" value="0.3" optional="true" /> <param name="max_bin" type="integer" label="Max Bin" value="256" optional="true" /> <param name="max_cat_threshold" type="integer" label="Max Cat Threshold" value="64" optional="true" /> <param name="max_cat_to_onehot" type="integer" label="Max Cat to Onehot" value="4" optional="true" /> <param name="max_delta_step" type="float" label="Max Delta Step" value="0" optional="true" /> <param name="max_depth" type="integer" label="Max Depth" value="6" optional="true" /> <param name="max_leaves" type="integer" label="Max Leaves" value="0" optional="true" /> <param name="min_child_weight" type="float" label="Min Child Weight" value="1" optional="true" /> <param name="monotone_constraints" type="text" label="Monotone Constraints" value="null" optional="true" /> <param name="multi_strategy" type="text" label="Multi Strategy" value="null" optional="true" /> <param name="n_estimators" type="integer" label="Number of Estimators" value="100" optional="true" /> <param name="num_parallel_tree" type="integer" label="Number of Parallel Trees" value="1" optional="true" /> <param name="random_state" type="integer" label="Random State" value="4131" /> <param name="reg_alpha" type="float" label="Reg Alpha" value="0" optional="true" /> <param name="reg_lambda" type="float" label="Reg Lambda" value="1" optional="true" /> <param name="sampling_method" type="select" label="Sampling Method" value="uniform"> <option value="uniform" selected="true" >Uniform</option> <option value="gradient_based">Gradient Based</option> </param> <param name="scale_pos_weight" type="float" label="Scale Pos Weight" value="1" optional="true" /> <param name="subsample" type="float" label="Subsample" value="1" optional="true" /> <param name="tree_method" type="select" label="Tree Method" value="auto"> <option value="auto" selected="true" >Auto</option> <option value="exact">Exact</option> <option value="approx">Approximate</option> <option value="hist">Histogram-based</option> <option value="gpu_hist">GPU Histogram-based</option> </param> <param name="validate_parameters" type="boolean" label="Validate Parameters" value="False" optional="true" /> <param name="verbosity" type="integer" label="Verbosity" value="1" /> </when> <when value="tune"> <param name="tune_param_file" type="data" format="txt" label="Training data" optional="false" argument= "--input_data"/> </when> </conditional> </when> </conditional> </when> <when value="lightgbm"> <conditional name="settings"> <param name="advanced" type="select" label="Use Advanced Settings to Create Model."> <option value="default" selected="true">Use Default Setup</option> <option value="settings">Use Advanced Setup</option> </param> <when value="default"> </when> <when value="settings"> <conditional name="settings"> <param name="advanced" type="select" label="Specify Advanced Parameters."> <option value="custom" selected="true">Create Model with Custom Parameters</option> <option value="tune">Tune Model</option> </param> <when value="custom"> <param name="boosting_type" type="select" label="Boosting Type"> <option value="gbdt">GBDT</option> <option value="dart">DART</option> <option value="goss">GOSS</option> <option value="rf">RF</option> </param> <param name="class_weight" type="text" value="null" label="Class Weight" optional="true"/> <param name="colsample_bytree" type="float" value="1.0" label="Column Sample by Tree"/> <param name="importance_type" type="select" label="Importance Type"> <option value="split" selected="true" >Split</option> <option value="gain">Gain</option> </param> <param name="learning_rate" type="float" value="0.1" label="Learning Rate"/> <param name="max_depth" type="integer" value="-1" label="Max Depth"/> <param name="min_child_samples" type="integer" value="20" label="Min Child Samples"/> <param name="min_child_weight" type="float" value="0.001" label="Min Child Weight"/> <param name="min_split_gain" type="float" value="0.0" label="Min Split Gain"/> <param name="n_estimators" type="integer" value="100" label="Number of Estimators"/> <param name="n_jobs" type="integer" value="-1" label="Number of Jobs"/> <param name="num_leaves" type="integer" value="31" label="Number of Leaves"/> <param name="objective" type="select" label="Objective"> <option selected='true' value="regression">regression</option> <option value="regression_l1">regression_l1</option> <option value="huber">huber</option> <option value="fair">fair</option> <option value="poisson">poisson</option> <option value="quantile">quantile</option> <option value="mape">mape</option> <option value="gamma">gamma</option> <option value="tweedie">tweedie</option> <option value="binary" >binary</option> <option value="multiclass">multiclass</option> <option value="multiclassova">multiclassova</option> <option value="cross_entropy">cross_entropy</option> <option value="cross_entropy_lambda">cross_entropy_lambda</option> <option value="lambdarank">lambdarank</option> <option value="rank_xendcg">rank_xendcg</option> </param> <param name="random_state" type="integer" value="4131" label="Random State"/> <param name="reg_alpha" type="float" value="0.0" label="Regularization Alpha"/> <param name="reg_lambda" type="float" value="0.0" label="Regularization Lambda"/> <param name="subsample" type="float" value="1.0" label="Subsample"/> <param name="subsample_for_bin" type="integer" value="200000" label="Subsample for Bin"/> <param name="subsample_freq" type="integer" value="0" label="Subsample Frequency"/> </when> <when value="tune"> <param name="tune_param_file" type="data" format="txt" label="Training data" optional="false" argument= "--input_data"/> </when> </conditional> </when> </conditional> </when> <when value="hdc"> <conditional name="settings"> <param name="advanced" type="select" label="Use Advanced Settings to Create Model."> <option value="default" selected="true">Use Default Setup</option> <option value="settings">Use Advanced Setup</option> </param> <when value="default"> </when> <when value="settings"> <conditional name="settings"> <param name="advanced" type="select" label="Specify Advanced Parameters."> <option value="custom" selected="true">Create Model with Custom Parameters</option> <option value="tune">Tune Model</option> </param> <when value="custom"> <param name="dimensionality" type="integer" label="Dimensionality" value="10000" /> <param name="levels" type="integer" label="Levels" value="100" /> <param name="retrain" type="integer" label="Retrain" value="0" /> </when> <when value="tune"> <param name="tune_param_file" type="data" format="txt" label="Tune parameter file" optional="false" argument= "--tune_para"/> </when> </conditional> </when> </conditional> </when> </conditional> </inputs> <outputs> <data name='output1' format='tabular' label="out.tsv" /> <data name='output2' format='html' label="repot.html" /> </outputs> <tests> <test> <param name="input1" value="test_count.tsv"/> <param name="input2" value="test_metadata.tsv"/> <conditional name="drop_columns"> <param name="advanced_setup" value="settings"/> <param name="columns_to_drop" value="1"/> </conditional> <param name="column_label" value="9"/> <param name="SelMLAlgo|MLAlgo" value="lr"/> <param name="setup|advanced_setup" value="default"/> <output name="output1" file="out.tsv" ftype="tabular"/> </test> <test> <param name="input1" value="test_count.tsv"/> <param name="input2" value="test_metadata.tsv"/> <conditional name="drop_columns"> <param name="advanced_setup" value="settings"/> <param name="columns_to_drop" value="1"/> </conditional> <param name="column_label" value="9"/> <param name="SelMLAlgo|MLAlgo" value="knn"/> <param name="setup|advanced_setup" value="default"/> <output name="output1" file="out.tsv" ftype="tabular"/> </test> <test> <param name="input1" value="test_count.tsv"/> <param name="input2" value="test_metadata.tsv"/> <conditional name="drop_columns"> <param name="advanced_setup" value="settings"/> <param name="columns_to_drop" value="1"/> </conditional> <param name="column_label" value="9"/> <param name="SelMLAlgo|MLAlgo" value="nb"/> <param name="setup|advanced_setup" value="default"/> <output name="output1" file="out.tsv" ftype="tabular"/> </test> <test> <param name="input1" value="test_count.tsv"/> <param name="input2" value="test_metadata.tsv"/> <conditional name="drop_columns"> <param name="advanced_setup" value="settings"/> <param name="columns_to_drop" value="1"/> </conditional> <param name="column_label" value="9"/> <param name="SelMLAlgo|MLAlgo" value="dt"/> <param name="setup|advanced_setup" value="default"/> <output name="output1" file="out.tsv" ftype="tabular"/> </test> <test> <param name="input1" value="test_count.tsv"/> <param name="input2" value="test_metadata.tsv"/> <conditional name="drop_columns"> <param name="advanced_setup" value="settings"/> <param name="columns_to_drop" value="1"/> </conditional> <param name="column_label" value="9"/> <param name="SelMLAlgo|MLAlgo" value="svm"/> <param name="setup|advanced_setup" value="default"/> <output name="output1" file="out.tsv" ftype="tabular"/> </test> <test> <param name="input1" value="test_count.tsv"/> <param name="input2" value="test_metadata.tsv"/> <conditional name="drop_columns"> <param name="advanced_setup" value="settings"/> <param name="columns_to_drop" value="1"/> </conditional> <param name="column_label" value="9"/> <param name="SelMLAlgo|MLAlgo" value="rf"/> <param name="setup|advanced_setup" value="default"/> <output name="output1" file="out.tsv" ftype="tabular"/> </test> <test> <param name="input1" value="test_count.tsv"/> <param name="input2" value="test_metadata.tsv"/> <conditional name="drop_columns"> <param name="advanced_setup" value="settings"/> <param name="columns_to_drop" value="1"/> </conditional> <param name="column_label" value="9"/> <param name="SelMLAlgo|MLAlgo" value="rbfsvm"/> <param name="setup|advanced_setup" value="default"/> <output name="output1" file="out.tsv" ftype="tabular"/> </test> <test> <param name="input1" value="test_count.tsv"/> <param name="input2" value="test_metadata.tsv"/> <conditional name="drop_columns"> <param name="advanced_setup" value="settings"/> <param name="columns_to_drop" value="1"/> </conditional> <param name="column_label" value="9"/> <param name="SelMLAlgo|MLAlgo" value="gbc"/> <param name="setup|advanced_setup" value="default"/> <output name="output1" file="out.tsv" ftype="tabular"/> </test> <test> <param name="input1" value="test_count.tsv"/> <param name="input2" value="test_metadata.tsv"/> <conditional name="drop_columns"> <param name="advanced_setup" value="settings"/> <param name="columns_to_drop" value="1"/> </conditional> <param name="column_label" value="9"/> <param name="SelMLAlgo|MLAlgo" value="xgboost"/> <param name="setup|advanced_setup" value="default"/> <output name="output1" file="out.tsv" ftype="tabular"/> </test> <test> <param name="input1" value="test_count.tsv"/> <param name="input2" value="test_metadata.tsv"/> <conditional name="drop_columns"> <param name="advanced_setup" value="settings"/> <param name="columns_to_drop" value="1"/> </conditional> <param name="column_label" value="9"/> <param name="SelMLAlgo|MLAlgo" value="lightgbm"/> <param name="setup|advanced_setup" value="default"/> <output name="output1" file="out.tsv" ftype="tabular"/> </test> <test> <param name="input1" value="test_count.tsv"/> <param name="input2" value="test_metadata.tsv"/> <conditional name="drop_columns"> <param name="advanced_setup" value="settings"/> <param name="columns_to_drop" value="1"/> </conditional> <param name="column_label" value="9"/> <param name="SelMLAlgo|MLAlgo" value="hdc"/> <param name="setup|advanced_setup" value="default"/> <output name="output1" file="out.tsv" ftype="tabular"/> </test> </tests> <help><![CDATA[ **Machine Learning Tool** This tool creates machine learning models using various classifiers and advanced configuration options for microbiome data analysis. **Inputs** - **Training Data**: A TSV file containing the features (columns) and samples (rows). - **Metadata**: A TSV file containing the sample metadata. - **Target Column (Class Label)**: The column in the metadata file that contains the class labels for classification. **Outputs** - **Model Results**: A TSV file containing the model performance metrics and predictions. - **Model Visualization**: An HTML file containing visualizations of the model results. **Features** - Support for multiple machine learning algorithms (Logistic Regression, K-Nearest Neighbors, Decision Tree, Random Forest, Gradient Boosting, XGBoost, LightGBM, Support Vector Machines, Naive Bayes, and HDC classifiers). - Advanced setup options for cross-validation, normalization, outlier removal, and dimensionality reduction. - Parameter tuning and custom parameter configuration for each algorithm. - Optional column dropping from training data. ]]></help> <citations> <citation type="bibtex"> @article{cumbo2023hdlib, title={hdlib: A Python library for designing Vector-Symbolic Architectures}, author={Cumbo, Fabio and Weitschek, Emanuel and Blankenberg, Daniel}, journal={Journal of Open Source Software}, volume={8}, number={89}, pages={5704}, year={2023} } </citation> <citation type="bibtex"> @article{cumbo2025feature, title={Feature selection with vector-symbolic architectures: a case study on microbial profiles of shotgun metagenomic samples of colorectal cancer}, author={Cumbo, Fabio and Truglia, Simone and Weitschek, Emanuel and Blankenberg, Daniel}, journal={Briefings in Bioinformatics}, volume={26}, number={2}, pages={bbaf177}, year={2025}, publisher={Oxford University Press} } </citation> </citations> </tool>