Quick start your ML model development with sklearn.
Define your own preprocessing pipeline and fit a model using preddefined estimators wth GridSearchCV as defined in estimators.py.
Currently defined regressors include
- ElasticNet
- RandomForest
- HistGradientBoost
Example
from sklearn.compose import ColumnTransformer
from sklearn_turbostart.estimators import create_estimators
from sklearn_turbostart.train import load_data, train_estimators
from sklearn.preprocessing import StandardScaler
# Load data
X_train, y_train, X_test, y_test = load_data("../data/data.csv")
# Preprocessing
preprocessing = ColumnTransformer(
[
(
"numeric_standard_scalar",
StandardScaler(),
["Feature 1", "Feature 2", "Feature 9"],
),
]
)
# Create sklearn Pipeline
estimators = create_estimators(preprocessing)
# Fit models
train_estimators(estimators, X_train, y_train, X_test, y_test)
After fitting, best estimator can be accessed using,
estimators[<estimator_name>: Optional["HistFradientBoost", "RandomForest", "ElasticNet"]].best_estimator_.
Check out Quickstarter.ipynb in notebooks/
React
Typescript
Django
Laravel
Facebook
Microsoft
Google
Alibaba
D3
Tencent