Title: An Automated Neural Network Approach for Credit Approval Prediction Using KNIME Analytics Platform.
Authors: Ihab M Aslim Samy S. Abu-Naser
Volume: 10
Issue: 7
Pages: 9-13
Publication Date: 2026/07/28
Abstract:
Credit risk assessment is a critical function for financial institutions to minimize default rates and optimize lending decisions. This research utilizes an Artificial Neural Network (ANN) trained via a Multi-Layer Perceptron (MLP) architecture within the KNIME Analytics Platform to automate and predict the credit approval process. Using the Credit Approval dataset from the UCI Machine Learning Repository, which contains a mix of categorical and continuous financial attributes, a robust preprocessing pipeline was implemented. The dataset was transformed, normalized, and partitioned into training and validation sets using dedicated KNIME nodes. The neural network successfully mapped non-linear relationships between applicant profiles and approval outcomes. The final model demonstrates high classification accuracy, proving that visual programming-based data analytics platforms can effectively support automated financial decision-making systems.