Paper: arXiv 2404.00013
Authors: Debarati Chakraborty, Ravi Ranjan
Abstract
This work focuses on designing a pipeline for the prediction of bankruptcy. The presence of missing values, high dimensional data, and highly class-imbalance databases are the major challenges in the said task. A new method for missing data imputation with granular semantics has been introduced here. The merits of granular computing have been explored here to define this method. The missing values have been predicted using the feature semantics and reliable observations in a low-dimensional space, in the granular space. The granules are formed around every missing entry, considering a few of the highly correlated features and most reliable closest observations to preserve the relevance and reliability, the context, of the database against the missing entries. An intergranular prediction is then carried out for the imputation within those contextual granules. That is, the contextual granules enable a small relevant fraction of the huge database to be used for imputation and overcome the need to access the entire database repetitively for each missing value. This method is then implemented and tested for the prediction of bankruptcy with the Polish Bankruptcy dataset. It provides an efficient solution for big and high-dimensional datasets even with large imputation rates. Then an AI-driven pipeline for bankruptcy prediction has been designed using the proposed granular semantic-based data filling method followed by the solutions to the issues like high dimensional dataset and high class-imbalance in the dataset. The rest of the pipeline consists of feature selection with the random forest for reducing dimensionality, data balancing with SMOTE, and prediction with six different popular classifiers including deep NN. All methods defined here have been experimentally verified with suitable comparative studies and proven to be effective on all the data sets captured over the five years.
Complexity vs Empirical Score
- Math Complexity: 4.5/10
- Empirical Rigor: 7.0/10
- Quadrant: Street Traders — empirical and applied, lower math complexity
Why this score: The paper uses moderately advanced concepts from granular computing and machine learning but focuses heavily on a practical pipeline (SMOTE, Random Forest, deep learning) applied to a specific real-world dataset (Polish Bankruptcy), with comparative experiments demonstrating effectiveness.
Research Flowchart
flowchart TD A["Research Goal: Improve Bankruptcy Prediction<br>with High-Dim, Imbalanced, Missing Data"] --> B["Input: Polish Bankruptcy Dataset"] B --> C["Method 1: Granular Semantic Imputation<br>(New Approach)"] C --> D["Form Contextual Granules<br>(Features + Reliable Observations)"] D --> E["Inter-Granular Prediction<br>(Low-dimensional Imputation)"] B --> F["Method 2: AI Prediction Pipeline"] E --> F F --> G["Feature Selection<br>(Random Forest)"] G --> H["Data Balancing<br>(SMOTE)"] H --> I["Classification<br>(Deep NN & Classifiers)"] I --> J["Findings: Effective for Big Data,<br>High Imputation Rates, & Class Imbalance"]
Evidence Card
What the authors claim, on what data, and who has checked it. Read from the paper's own text by a language model: every quote is verbatim, nothing was re-run. How to read this card
Proposes
The paper introduces a granular semantic missing‑data imputation method and an AI‑driven pipeline for bankruptcy prediction.
Source sentence
A new method for missing data imputation with granular semantics has been introduced here.
Assumes 2
- All features are assumed to be numerical. data
Source sentence
In this work, we assumed all features to be numerical.
- Granules are formed in a low‑dimensional space where the granule dimension δ is much smaller than the total number of features d. model
Source sentence
Let the granule formed around the missing values be of dimension δ, where δ << d.
Data
- Polish Bankruptcy dataset other unknown
Source sentence
12789 observations from the Polish Bankruptcy data set throughout all five years are used to demonstrate the effectiveness of feature reduction and data balancing in the pipeline.
Costs
Not applicable
Finds 3
- The end‑to‑end pipeline achieves accuracies greater than 90% for bankruptcy prediction in most cases. predictive-accuracy
Source sentence
The end-to-end pipeline designed here results in accuracies more than 90% for the prediction of bankruptcy in most cases.
- The Granular Semantic imputation method remains consistent with low impurity (<10%) and outperforms other methods as impurity increases. performance
Source sentence
It can be observed from the figure that the performance of the proposed Granular Semantic method is almost consistent and as good as the other methods with low impurity (< 10%). Once the impurity increases, the proposed method performed better in all cases compared to the other methods.
- The new contextual‑granule imputation achieves higher or comparable accuracy to autoencoder‑based estimators. predictive-accuracy
Source sentence
The newly formulated data imputation technique with contextual granule has been compared with three other popular methods, and resulted in higher or almost equal accuracy even compared to autoencoder-based estimators.
Checked by others
No other paper in the archive has been found to support or contradict these findings (compared with 13 related papers). Coverage is partial, so this is neither replication nor refutation.
Code
No repository by the authors found in the paper.
Extracted 2026-11-14 · same card as JSON · A language model chose these items: check the source sentence before relying on one.
Cited References (47)
- Real-time segmentation of on-line handwritten arabic script (2014)
- Fast classification of handwritten on-line Arabic characters (2014) — link
- Estimate and Replace: A Novel Approach to Integrating Deep Neural Networks with Existing Applications (2018) — link
- Sales Gets a Machine-Learning Makeover (2016)
- Analysis of financial credit risk using machine learning (2018) — link
- Machine learning predictivity applied to consumer creditworthiness (2020)
- Machine learning in banking risk management: A literature review (2019)
- Deep learning models for bankruptcy prediction using textual disclosures (2019)
and 39 more
- Forecasting bankruptcy in the wood industry (2021)
- Bankruptcy prediction using deep learning approach based on borderline SMOTE (2020)
- Ensemble boosted trees with synthetic features generation in application to bankruptcy prediction (2016)
- A profit-driven Artificial Neural Network (ANN) with applications to fraud detection and direct marketing (2016)
- Financial credit risk assessment: a recent review (2016)
- Bankruptcy prediction using machine learning (2017)
- Extreme Gradient Boosting Method in the Prediction of Company Bankruptcy (2019)
- Bankruptcy prediction in banks and firms via statistical and intelligent techniques–A review (2007)
- Creation Bankruptcy Prediction Model with Using Ohlson and Shirata Models (2013)
- A review of bankruptcy prediction studies: 1930 to present (2007)
- Using News Articles and Financial Data to predict the likelihood of bankruptcy (2020) — link
- Bankruptcy prediction on the base of the unbalanced data using multi-objective selection of classifiers (2021)
- Generative Imputation and Stochastic Prediction (2022)
- Systematic Review of Using Machine Learning in Imputing Missing Values (2022)
- Missing value imputation affects the performance of machine learning: A review and analysis of the literature (2010–2021) (2021)
- Multivariate time series imputation for energy data using neural networks (2023)
- Improved Random Forest for Classification (2018)
- SMOTE: synthetic minority over-sampling technique (2002)
- Polish companies bankruptcy data (2016)
- Impacts of Fractional Hot-Deck Imputation on Learning and Prediction of Engineering Data (2020)
- An Experimental Survey of Missing Data Imputation Algorithms (2023)
- Missing value imputation in food composition data with denoising autoencoders (2022)
- Imputation of Missing Values in Time Series Using an Adaptive-Learned Median-Filled Deep Autoencoder (2023)
- Regression multiple imputation for missing data analysis (2020)
- Multiple imputation by chained equations: what is it and how does it work? (2011)
- Review of bankruptcy prediction using machine learning and deep learning techniques (2019)
- Application of Artificial Intelligence in the Supply Chain Finance (2023)
- Machine Learning for Financial Risk Management: A Survey (2020)
- Deep Learning for Intelligent Assessment of Financial Investment Risk Prediction (2022)
- Granular Video Computing with Rough Sets, Deep Learning and in IoT (2021)
- Toward a Theory of Fuzzy Information Granulation and Its Centrality in Human Reasoning and Fuzzy Logic (1997)
- A Granular Computing-Driving Hesitant Fuzzy Linguistic Method for Supporting Large-Scale Group Decision Making (2022)
- The Long-Term Prediction of Time Series: A Granular Computing-Based Design Approach (2022)
- Event prediction with rough-fuzzy sets (2023)
- Q-rough sets, flicker modeling and unsupervised fire threat quantification from videos (2022)
- Statistical analysis with missing data (2019)
- Deep learning versus conventional methods for missing data imputation: A review and comparative study (2023)
- Missing Value Imputation Based on Gaussian Mixture Model for the Internet of Things (2015)
- Imputing missing data for gene expression arrays (1999)
0 of 47 cited works are in the corpus; resolved via arXiv source bundles, Semantic Scholar and OpenAlex. Corpus-wide: 2596 linked papers.