Export scikit-learn model files to JSON for sharing or deploying predictive models with peace of mind.
Other methods for exporting scikit-learn models require Pickle or Joblib (based on Pickle). Serializing model files with Pickle provides a simple attack vector for malicious users-- they give an attacker the ability to execute arbitrary code wherever the file is deserialized. For an example see: https://www.smartfile.com/blog/python-pickle-security-problems-and-solutions/.
sklearn-json is a safe and transparent solution for exporting scikit-learn model files.
Export model files to 100% JSON which cannot execute code on deserialization.
Model files are serialized in JSON (i.e., not binary), so you have the ability to see exactly what's inside.
sklearn-json makes exporting model files to JSON simple.
pip install sklearn-json
import sklearn_json as skljson
from sklearn.ensemble import RandomForestClassifier
model = RandomForestClassifier(n_estimators=10, max_depth=5, random_state=0).fit(X, y)
skljson.to_json(model, file_name)
deserialized_model = skljson.from_json(file_name)
deserialized_model.predict(X)
The list of supported models is rapidly growing. If you have a request for a model or feature, please reach out to [email protected].
sklearn-json requires scikit-learn >= 0.21.3.
-
Classification
sklearn.linear_model.LogisticRegression
sklearn.linear_model.Perceptron
sklearn.discriminant_analysis.LinearDiscriminantAnalysis
sklearn.discriminant_analysis.QuadraticDiscriminantAnalysis
sklearn.svm.SVC
sklearn.naive_bayes.GaussianNB
sklearn.naive_bayes.MultinomialNB
sklearn.naive_bayes.ComplementNB
sklearn.naive_bayes.BernoulliNB
sklearn.tree.DecisionTreeClassifier
sklearn.ensemble.RandomForestClassifier
sklearn.ensemble.GradientBoostingClassifier
sklearn.neural_network.MLPClassifier
-
Regression
sklearn.linear_model.LinearRegression
sklearn.linear_model.Ridge
sklearn.linear_model.Lasso
sklearn.svm.SVR
sklearn.tree.DecisionTreeRegressor
sklearn.ensemble.RandomForestRegressor
sklearn.ensemble.GradientBoostingRegressor
sklearn.neural_network.MLPRegressor