A microservices-based Streaming and Batch data processing in Cloud Foundry and Kubernetes
orchestration
stream-processing
cloud-native
microservices-architecture
predictive-analytics
batch-processing
datapipelines
-
Updated
Jul 24, 2021 - Java
We compute variable importance by calculation pearson correlation between scores and targed encoded variables:
https://github.com/PythonPredictions/cobra/blob/0133435194ac63f771e6d165c23ea34233dd0a94/cobra/model_building/models.py#L144-L150
It'd nice to choose different correlation (like kendall)?
Because perason assumes normality, but these two are not normally distributed.
https://data