Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

KAXAI: An Integrated Environment for Knowledge Analysis and Explainable AI

Saikat Barua, Sifat Momen

TL;DR

KAXAI addresses the challenge of making machine learning more accessible by integrating AutoML, Explainable AI, and synthetic data generation into a single, user-friendly environment. The framework combines two novel classifiers, Logistic Regression Forest and Support Vector Tree, with a model-dependent interpreter called MEDLEY to provide local explanations, while leveraging GANs and LLMs for dataset augmentation. The paper reports competitive accuracy on diabetes and survey datasets, demonstrates robust interpretation capabilities, and showcases multiple data-generation strategies with performance and statistical analyses. The work aims to reduce integration friction, improve model transparency, and expand usable data through synthetic generation, offering practical benefits for practitioners and researchers working with tabular data.

Abstract

In order to fully harness the potential of machine learning, it is crucial to establish a system that renders the field more accessible and less daunting for individuals who may not possess a comprehensive understanding of its intricacies. The paper describes the design of a system that integrates AutoML, XAI, and synthetic data generation to provide a great UX design for users. The system allows users to navigate and harness the power of machine learning while abstracting its complexities and providing high usability. The paper proposes two novel classifiers, Logistic Regression Forest and Support Vector Tree, for enhanced model performance, achieving 96\% accuracy on a diabetes dataset and 93\% on a survey dataset. The paper also introduces a model-dependent local interpreter called MEDLEY and evaluates its interpretation against LIME, Greedy, and Parzen. Additionally, the paper introduces LLM-based synthetic data generation, library-based data generation, and enhancing the original dataset with GAN. The findings on synthetic data suggest that enhancing the original dataset with GAN is the most reliable way to generate synthetic data, as evidenced by KS tests, standard deviation, and feature importance. The authors also found that GAN works best for quantitative datasets.

KAXAI: An Integrated Environment for Knowledge Analysis and Explainable AI

TL;DR

KAXAI addresses the challenge of making machine learning more accessible by integrating AutoML, Explainable AI, and synthetic data generation into a single, user-friendly environment. The framework combines two novel classifiers, Logistic Regression Forest and Support Vector Tree, with a model-dependent interpreter called MEDLEY to provide local explanations, while leveraging GANs and LLMs for dataset augmentation. The paper reports competitive accuracy on diabetes and survey datasets, demonstrates robust interpretation capabilities, and showcases multiple data-generation strategies with performance and statistical analyses. The work aims to reduce integration friction, improve model transparency, and expand usable data through synthetic generation, offering practical benefits for practitioners and researchers working with tabular data.

Abstract

In order to fully harness the potential of machine learning, it is crucial to establish a system that renders the field more accessible and less daunting for individuals who may not possess a comprehensive understanding of its intricacies. The paper describes the design of a system that integrates AutoML, XAI, and synthetic data generation to provide a great UX design for users. The system allows users to navigate and harness the power of machine learning while abstracting its complexities and providing high usability. The paper proposes two novel classifiers, Logistic Regression Forest and Support Vector Tree, for enhanced model performance, achieving 96\% accuracy on a diabetes dataset and 93\% on a survey dataset. The paper also introduces a model-dependent local interpreter called MEDLEY and evaluates its interpretation against LIME, Greedy, and Parzen. Additionally, the paper introduces LLM-based synthetic data generation, library-based data generation, and enhancing the original dataset with GAN. The findings on synthetic data suggest that enhancing the original dataset with GAN is the most reliable way to generate synthetic data, as evidenced by KS tests, standard deviation, and feature importance. The authors also found that GAN works best for quantitative datasets.
Paper Structure (47 sections, 20 figures, 13 tables, 6 algorithms)

This paper contains 47 sections, 20 figures, 13 tables, 6 algorithms.

Table of Contents

  1. Introduction
  2. Literature Review
  3. Related Work on AutoML
  4. Related Work on XAI
  5. Related Work on Synthetic Data Generation
  6. Methodology
  7. Automated Machine Learning
  8. Data Preparation
  9. Data Visualization
  10. Model Selection
  11. List of Models
  12. Hyperparameter optimization
  13. Model Training
  14. Model Evaluation
  15. Feature and Drop Column Importance
  16. ...and 32 more sections

Figures (20)

  • Figure 1: KAXAI Methodolgy
  • Figure 2: Logistic Regression Forest Algo Visualization
  • Figure 3: Support Vector Tree Algo Visualization
  • Figure 4: MEDLEY Algo Visualization
  • Figure 5: Heatmap of Survey and Diabetes Dataset
  • ...and 15 more figures