algo_name

Description

Specifies the name of the algorithm used for training. Identifying the algorithm by name allows for precise documentation and easier replication of model training processes. This setting is fundamental for tracing model performance back to the algorithmic choices made during setup.

Options

• arima

• catboost

• decisiontree

• dummy

• extratrees

• identity

• knn

• lightgm

• logistic

• prophet

• randomforest

• sarima

• svm

• xgboost

Steps impacted

training

Example

date_col: Date_Chargement
algo_type: regression
algo_name: xgboost
algorithm_parameters:
  n_estimators: 150
  max_depth: 4