openscoring's profile picture. We bring predictive models to enterprise Java applications

Openscoring

@openscoring

We bring predictive models to enterprise Java applications

JPMML-Transpiler has been upgraded from 1.2.X to 1.3.X! During modularization, the core module was renamed from 'org.jpmml:jpmml-transpiler' to 'org.jpmml:pmml-transpiler'.


SkLearn2PMML 0.83.0 approves this. All data structures unchanged, except for minor tweaks around GB and HistGB loss functions (again!)

scikit-learn 1.1 is out! What's new? You can check the release highlights there: bit.ly/3yFnT7Q pip install -U scikit-learn or conda install -c conda-forge scikit-learn #sklearn #ML #Datascience #opensource #Python

scikit_learn's tweet image. scikit-learn 1.1 is out!
What's new? You can check the release highlights there: bit.ly/3yFnT7Q

pip install -U scikit-learn

or 

conda install -c conda-forge scikit-learn 

#sklearn #ML #Datascience #opensource #Python


New project for exporting fitted #Spark pipeline models to #PMML documents. Interoperable with #PySpark: github.com/jpmml/jpmml-sp…


Using #Spark pipeline models for real-time prediction: the #REST web service approach openscoring.io/blog/2016/07/0…


@bdy11n Export using JPMML-XGBoost and interpret like any other gradient boosting model - tree paths, vars contributing to each path


Openscoring reposted

Architecting a Machine Learning System for Risk twistopayments.github.io/machine-learni… Thanks @airbnbnerds for the inspiration! #MachineLearning


.@clarler Several users in Netflix. Check out this C6Stephens' mod for multiple-model evaluation use case: github.com/c6stephens/ope…


Openscoring reposted

@openscoring great work releasing a pmml extractor for #scikit-learn! #python #datascience


Loading...

Something went wrong.


Something went wrong.