I have built a repository for bagging pipeline with Node-red front end control. Bagging is an Ensemble learning technique that allows partition of a whole data set into different sub-set of dataset for seperate training processes and aggregation of a final prediction result given different balloting strategy. Bagging is a useful technique normally seen in federated learning for many industrial applications such as IT security and medical services where privacy is an issue. Our bagging pipeline has data separation, decision tree training and balloting steps. Each is wrapped into pipeline components. A complete pipeline docker file is constructed and a yaml file is conbined within a Node-red custom node for easy deployment. A complete file set and readme can be seen in the following repository:
https://github.com/Lance410247/kubeflow-node-red.
I have built a repository for bagging pipeline with Node-red front end control. Bagging is an Ensemble learning technique that allows partition of a whole data set into different sub-set of dataset for seperate training processes and aggregation of a final prediction result given different balloting strategy. Bagging is a useful technique normally seen in federated learning for many industrial applications such as IT security and medical services where privacy is an issue. Our bagging pipeline has data separation, decision tree training and balloting steps. Each is wrapped into pipeline components. A complete pipeline docker file is constructed and a yaml file is conbined within a Node-red custom node for easy deployment. A complete file set and readme can be seen in the following repository: https://github.com/Lance410247/kubeflow-node-red.