Algorithm
Penalty
C

Algorithm		Split quality criterion	Gini
Number of trees		Use bootstrap	Yes
Max trees depth		Feature sampling strategy
Min samples per leaf		Used features		Used features
Min samples to split

Algorithm		Split quality criterion	Gini
Number of trees		Use bootstrap	Yes
Max trees depth		Feature sampling strategy
Min samples per leaf		Used features		Used features
Min samples to split

Algorithm		Split quality criterion	MSE
Number of trees		Use bootstrap	Yes
Max trees depth		Feature sampling strategy
Min samples per leaf		Used features		Used features
Min samples to split

Algorithm		Split quality criterion	MSE
Number of trees		Use bootstrap	Yes
Max trees depth		Feature sampling strategy
Min samples per leaf		Used features		Used features
Min samples to split

Algorithm	SVM classifier
Kernel
Kernel coef (gamma)
C
Independent kernel term
Stopping tolerance
Max iterations

Algorithm	SVM regressor
Kernel
Kernel coef (gamma)
C
Independent kernel term
Stopping tolerance
Max iterations

Algorithm	SGD
Loss function
Epsilon
Penalty
L1 ratio
Stopping tolerance
Max iterations
Actual iterations

Algorithm	Decision tree classifier
Max tree depth
Split criterion
Min samples per leaf
Splitter

Algorithm	Ridge regression (L2)
Alpha
Alpha

Algorithm	Lasso regression (L1)
Alpha
Alpha

Algorithm	Ordinary Least Squares regression

Algorithm	Logistic Regression (MLLib)
Max iterations
Lambda (regularization param)
Alpha (Elastic net param)

Algorithm	Linear Regression (MLLib)
Max iterations
Lambda (regularization param)
Alpha (Elastic net param)

Algorithm	Random Forest (MLLib)
Number of trees
Maximum depth of tree
Step size
Feature subset strategy
Impurity
Maximum number of bins
Maximum memory
Check point interval
Cache node IDs
Minimum information gain
Minimum instance per node
Subsampling rate
Subsampling seed

Algorithm	Gradient Boosted Trees (MLLib)
Number of trees
Maximum depth of tree
Step size
Impurity
Maximum number of bins
Maximum memory
Check point interval
Cache node IDs
Minimum information gain
Minimum instance per node
Subsampling rate
Step size
Subsampling seed

Algorithm	Decision Tree (MLLib)
Maximum depth of tree
Ids of cache nodes
Checkpoint interval
Maximum number of bins
Maximum memory
Minimum information gain
Minimum instance per node

Algorithm	Naive Bayes (MLLib)
Lambda

Algorithm	XGBoost
Booster
Objective
Actual number of trees
Max trees depth
Eta (learning rate)
Max delta step
Alpha (L1 regularization)
Lambda (L2 regularization)
Gamma (Min loss reduction to split a leaf)
Min sum of instance weight in a child
Subsample ratio of the training instance
Columns subsample ratio for trees
Columns subsample ratio for splits / levels
Balancing of positive and negative weights
Value treated as missing		NaN
Tweedie variance power

Algorithm	Gradient Boosted Trees ({{(postTrain.algorithm === 'GBT_CLASSIFICATION') ? 'Classification' : 'Regression'}})
Loss	Deviance Exponential Least Square Least Absolute Deviation Huber
Feature sampling strategy	Default Square root Logarithm Fixed number Fixed proportion
Number of boosting stages
Eta (learning rate)
Max trees depth
Minimum samples per leaf

Algorithm	XGBoost
Booster
Actual number of trees
Max trees depth
Eta (learning rate)
Alpha (L1 regularization)
Lambda (L2 regularization)
Gamma (Min loss reduction to split a leaf)
Min sum of instance weight in a child
Subsample ratio of the training instance
Fraction of columns in each tree
Value treated as missing		NaN

Algorithm	LightGBM
Booster	{{ postTrain.lightgbm.boosting_type }}
Actual number of trees	{{ postTrain.lightgbm.n_estimators }}
Maximum number of leaves	{{ postTrain.lightgbm.num_leaves }}
Learning rate	{{ postTrain.lightgbm.learning_rate }}
Alpha (L1 regularization)	{{ postTrain.lightgbm.reg_alpha }}
Lambda (L2 regularization)	{{ postTrain.lightgbm.reg_lambda }}
Minimal gain to perform a split on a leaf	{{ postTrain.lightgbm.min_split_gain }}
Minimum leaf samples	{{ postTrain.lightgbm.min_child_samples }}
Min sum of instance weight in a child	{{ postTrain.lightgbm.min_child_weight }}
Subsample ratio of the training instance	{{ postTrain.lightgbm.subsample }}
Columns subsample ratio for trees	{{ postTrain.lightgbm.colsample_bytree }}

Algorithm	Logistic regression (Vertica)
Max number of iterations

Algorithm	Decision Tree
Maximum depth
Min. samples per leaf
Split strategy	Random Best

Algorithm	Neural Network
Activation	ReLU Identity Logistic Hyperbolic Tangent
Alpha
Max iterations
Convergence tolerance
Early stopping
Validation fraction
Solver	ADAM Stochastic Gradient Descent LBFGS
Shuffle data
Intial Learning Rate
Automatic batching
Batch size
beta_1
beta_2
epsilon
Learning rate annealing	Constant Inverse scaling Adaptive
power_t
Momentum
Use Nesterov momentum

Algorithm	KNN
Neighbor finding algorithm	Automatic KD Tree Ball Tree Brute force
K
Distance weighting
Leaf size
p

Algorithm	LASSO-LARS
Max number of features

Algorithm	Linear regression (Vertica)
Max number of iterations

Algorithm	Neural Network built with Keras

Algorithm	Deep Neural Network
Hidden layers
Units per layer
Learning rate
Early stopping
Early stopping patience
Early stopping threshold
Batch size
Number of epochs
Dropout
Lambda (L2 regularization)
Alpha (L1 regularization)

Algorithm

Algorithm	Naive seasonal forecasting
Season length

Algorithm	Auto ARIMA
Season length
Information criterion
Solver
Stationary
Maximum iterations
Unit root test
Seasonal unit root test
Forecast quantiles

Algorithm	ARIMA
p
d
q
p
P
D
Q
s
Trend
Trend offset
Enforce stationarity
Enforce invertibility
Concentrate scale

Algorithm	ETS
Short name
Trend
Damped trend
Error
Seasonal
Season length
Seed

Algorithm
Season length
Seasonal smoother length
Trend smoother length
Low pass length
Degree of seasonal LOESS
Degree of trend LOESS
Degree of low pass LOESS
Seasonal jump
Trend jump
Low pass jump
Forecast quantiles

Algorithm
Changepoint prior scale
Growth
Floor
Capacity
Seasonality prior scale
Seasonality mode
Yearly seasonality
Weekly seasonality
Daily seasonality
External features prior scale
Changepoint range
Number of changepoints
Seed

Algorithm	GluonTS NPTS forecaster
Context length
Kernel type
Exponential kernel weights
Seasonal model
Use default time features
Feature scale
Forecast quantiles

Algorithm	GluonTS DeepAR
Context length
Output distribution	Student's t	Negative binomial
Use identifiers as features
Nb. RNN layers
Nb. cells per layer
Dropout rate
Nb. batches per epoch
Learning rate
Weight decay
Patience
Batch size
Epochs
Forecast quantiles

Algorithm	GluonTS Feed forward neural network
Context length
Output distribution	Student's t	Negative binomial
Batch normalization
Hidden layer sizes
Nb. batches per epoch
Learning rate
Weight decay
Batch size
Epochs
Forecast quantiles

Algorithm	GluonTS Feed forward neural network
Context length
Output distribution	Student's t	Gaussian	Negative binomial
Batch normalization
Mean scaling
Hidden layer sizes
Nb. batches per epoch
Learning rate
Batch size
Epochs
Forecast quantiles

Algorithm	GluonTS DeepAR
Context length
Output distribution	Student's t	Gaussian	Negative binomial
Use identifiers as features
Nb. RNN layers
Nb. cells per layer
Cell type
Dropout cell type	Zoneout	RNN zoneout	Variational dropout	Variational zoneout
Dropout rate
α
β
Nb. batches per epoch
Learning rate
Batch size
Epochs
Forecast quantiles

Algorithm	GluonTS Transformer
Context length
Output distribution	Student's t	Gaussian	Negative binomial
Use identifiers as features
Transformer network dimension
Hidden layer dimension scale
Nb. heads in multi-head attention
Nb. batches per epoch
Learning rate
Batch size
Epochs
Forecast quantiles

Algorithm	GluonTS Multi-horizon quantile CNN
Context length
Use identifiers as features
MLP layer sizes
Nb. channels
Convolution dilations
Kernel sizes
Nb. batches per epoch
Learning rate
Batch size
Epochs
Forecast quantiles

Algorithm	Causal forest	Split quality criterion	{{ postTrain.causal_forest_params.criterion }}
Number of trees
Max trees depth		Feature sampling strategy
Min samples per leaf		Used features		Used features

Algorithm	{{postTrain.algorithm}}