Icon

Building a Credit Scoring Model

Credit Scoring

This KNIME workflow focuses on creating a credit scoring model based on historical data. As with all data mining modeling activities, it is unclear in advance which analytic method is most suitable. This workflow therefore uses three different methods simultaneously – Decision Trees, Neural Networking and SVM – then automatically determines which model is most accurate and writes that model out for further use.
This workflow manipulates the data so it is suitable for a variety of modeling techniques by converting nominals to numerics. The data was enhanced so that understandable labels are used. It uses metanodes to “package” each technique suitable for reuse. Each Model uses a Test / Learn and cross validated process to ensure accuracy. The workflow writes out the model in the official PMML format, so that other applications can use the model.

Credit ScoringCredit scoring is a technique used to determine whether or not to extend credit (and if so, how much) to a borrower. This workflow illustrates how to create and choose a credit scoring modelbased on both historical data and on the application of different machine learning algorithms.Task Create a credit scoring model based on historical data. Select the best machine learning algorithm to be applied. Use cross-validation to evaluate model performance.A use case is described at URL: https://www.knime.org/knime-applications/credit-scoring Data ReadingThe data are GermanCredit data, includingcredit status,demographic data, andcustomer history. Thefile is located inTheData/Credit Pre-processingLearners such as neuralnetwork or SVM can onlyhandle numeric attributes.Nominal columns areconverted into numericalcolumns. Model Training and Evaluation1)The following algorithms are trained andevaluated with cross-validation: - Neural Network - SVM - Decision Tree2) Double-click on the metanode to see thesubworkflow Model SelectionAll results, i.e. accuracies and respectivemodels, are combined in one single table.Rows are then sorted by descendingaccuracy and only first row (best performingmodel) is kept. Bar Chart Save the Model - Convert the model cell back to PMML - Save the model. KNIME Analytics Platform writes out themodel in the official PMML format, so thatother applications can use the model. Try this:1) Choose your own algorithm and concatenate itwith the other algorithms. Check if your algorithmperforms better than the others.2) Change the aggregation method to "Sum" or"Average" in the view, to see the accuracies. Sort by AccuracyPick up the best modelReadingcredit scoring dataset Train and Cross Validatea Decision Tree Train and Cross Validatea Neural Network Train and CrossValidate a SVM Concatenate(Optional in) Sorter Row Filter Cell To PMML PMML Writer File Reader Category To Number JavaScriptBar Chart Credit ScoringCredit scoring is a technique used to determine whether or not to extend credit (and if so, how much) to a borrower. This workflow illustrates how to create and choose a credit scoring modelbased on both historical data and on the application of different machine learning algorithms.Task Create a credit scoring model based on historical data. Select the best machine learning algorithm to be applied. Use cross-validation to evaluate model performance.A use case is described at URL: https://www.knime.org/knime-applications/credit-scoring Data ReadingThe data are GermanCredit data, includingcredit status,demographic data, andcustomer history. Thefile is located inTheData/Credit Pre-processingLearners such as neuralnetwork or SVM can onlyhandle numeric attributes.Nominal columns areconverted into numericalcolumns. Model Training and Evaluation1)The following algorithms are trained andevaluated with cross-validation: - Neural Network - SVM - Decision Tree2) Double-click on the metanode to see thesubworkflow Model SelectionAll results, i.e. accuracies and respectivemodels, are combined in one single table.Rows are then sorted by descendingaccuracy and only first row (best performingmodel) is kept. Bar Chart Save the Model - Convert the model cell back to PMML - Save the model. KNIME Analytics Platform writes out themodel in the official PMML format, so thatother applications can use the model. Try this:1) Choose your own algorithm and concatenate itwith the other algorithms. Check if your algorithmperforms better than the others.2) Change the aggregation method to "Sum" or"Average" in the view, to see the accuracies. Sort by AccuracyPick up the best modelReadingcredit scoring dataset Train and Cross Validatea Decision Tree Train and Cross Validatea Neural Network Train and CrossValidate a SVM Concatenate(Optional in) Sorter Row Filter Cell To PMML PMML Writer File Reader Category To Number JavaScriptBar Chart

Nodes

Extensions

Links

Download

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

Download Workflow
On NodePit since: 2019-10-03
Last update: 2024-11-21
Created with KNIME version: v3.5.0
Tags: Credit Scoring Credit Rating Customer Risk Decision Tree Neural Network SVM

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

Try NodePit Runner!