Icon

JKISeason3-18

Description: You work as a researcher creating models to identify whether a breasttumor is benign or malign, based on anonymized patient data. Besides obtaining aclassifier that works very well for both benign and malign cases, you must be able toexplain how different feature values impact your results. Experiment with LIME andvisualization techniques to explain your predictions and make your research moretransparent. Hint: Learn more about this problem's data attributes here. Used a minimal neural network - as inputfor LIME analysis DataNode 583#2 = # 4478collectsexplanationsNode 592top input : test set instance rows to be explainedbottom input : test set distributiontop: 90% train & validation setbottom: 10% testtop: 90% train & validation setbottom: 10% testNode 596Node 597Node 598training a local GLMfor each input instance to generate a Local Inter. Model-agn. ExplanationNode 612Node 613Node 614Node 615Node 616Node 617Node 618Node 620Node 621Node 622Node 623Node 624Node 625Node 626Node 627Node 628Node 630Node 631Node 632Node 640Node 647Node 649Node 650Node 651Node 653Node 658CSV Reader Normalizer Equal Size Sampling Loop End Chunk Loop Start LIME Loop Start Partitioning Partitioning Column Filter Loop End Flag DuplicateSample Code Number Compute LIME Column Filter Keras NetworkLearner Keras Dense Layer Keras Input Layer Keras NetworkExecutor Keras Dense Layer Rule Engine String to Number Row Splitter Column Appender Column Filter Column Filter Column Appender Unpivot Group Loop Start Row Splitter Joiner Column Filter Pivot Loop End Rule Engine Column Renamer Joiner Unpivot Column Renamer String to Number Round Double DataViz Description: You work as a researcher creating models to identify whether a breasttumor is benign or malign, based on anonymized patient data. Besides obtaining aclassifier that works very well for both benign and malign cases, you must be able toexplain how different feature values impact your results. Experiment with LIME andvisualization techniques to explain your predictions and make your research moretransparent. Hint: Learn more about this problem's data attributes here. Used a minimal neural network - as inputfor LIME analysis DataNode 583#2 = # 4478collectsexplanationsNode 592top input : test set instance rows to be explainedbottom input : test set distributiontop: 90% train & validation setbottom: 10% testtop: 90% train & validation setbottom: 10% testNode 596Node 597Node 598training a local GLMfor each input instance to generate a Local Inter. Model-agn. ExplanationNode 612Node 613Node 614Node 615Node 616Node 617Node 618Node 620Node 621Node 622Node 623Node 624Node 625Node 626Node 627Node 628Node 630Node 631Node 632Node 640Node 647Node 649Node 650Node 651Node 653Node 658CSV Reader Normalizer Equal Size Sampling Loop End Chunk Loop Start LIME Loop Start Partitioning Partitioning Column Filter Loop End Flag DuplicateSample Code Number Compute LIME Column Filter Keras NetworkLearner Keras Dense Layer Keras Input Layer Keras NetworkExecutor Keras Dense Layer Rule Engine String to Number Row Splitter Column Appender Column Filter Column Filter Column Appender Unpivot Group Loop Start Row Splitter Joiner Column Filter Pivot Loop End Rule Engine Column Renamer Joiner Unpivot Column Renamer String to Number Round Double DataViz

Nodes

Extensions

Links

Download

To use this workflow in KNIME, download it from the below URL and open it in KNIME:

Download Workflow
Created by: Adrian Nembach
Created at: 2019-06-21
On NodePit since: 2024-09-17
Last update: 2024-09-19
Created with KNIME version: v5.2.5
Tags: JKISeason3-18

Deploy, schedule, execute, and monitor your KNIME workflows locally, in the cloud or on-premises – with our brand new NodePit Runner.

Try NodePit Runner!