Missing Data Imputation With Granular Semantics and AI-driven Pipeline for Bankruptcy Prediction

• DOI: 10.48550/arxiv.2404.00013
• Publication Date: 2024-03-15
• Authors: Debarati Bhunia C...; Ravi Ranjan
• Keywords: Imputation (statistics); Bankruptcy Prediction

This work focuses on designing a pipeline for the prediction of bankruptcy. The presence missing values, high dimensional data, and highly class-imbalance databases are major challenges in said task. A new method data imputation with granular semantics has been introduced here. merits computing have explored here to define this method. values predicted using feature reliable observations low-dimensional space, space. granules formed around every entry, considering few correlated features most closest preserve relevance reliability, context, database against entries. An intergranular is then carried out within those contextual granules. That is, enable small relevant fraction huge be used overcome need access entire repetitively each value. implemented tested bankruptcy Polish Bankruptcy dataset. It provides an efficient solution big high-dimensional datasets even large rates. Then AI-driven designed proposed semantic-based filling followed by solutions issues like dataset rest consists selection random forest reducing dimensionality, balancing SMOTE, six different popular classifiers including deep NN. All methods defined experimentally verified suitable comparative studies proven effective all sets captured over five years.