Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

A multi-model approach using XAI and anomaly detection to predict asteroid hazards

Amit Kumar Mondal, Nafisha Aslam, Prasenjit Maji, Hemanta Kumar Mondal

TL;DR

The paper addresses the hazard prediction of near-Earth asteroids (PHAs) by proposing a hybrid ML-DL framework that fuses textual orbital/numerical data and image data from telescope observations. It integrates Explainable AI (SHAP, LIME, Grad-CAM) and anomaly detection (Isolation Forest, One-Class SVM, Autoencoders) into a weighted ensemble of models (textual: LR, DT, RF, SVM, KNN, Naive Bayes, XGBoost; visual: CNN, EfficientNetB0, VGG16) to deliver real-time hazard predictions and alerts. Advanced preprocessing (imputation, outlier removal, scaling, PCA, RFE, augmentation) and a real-time alert system (API access, dashboards, automatic notifications) are central to robustness and timeliness. The results show top accuracies up to 99% for text-based classifications and 98–99% for image-based classifications, with XAI and anomaly detection enhancing interpretability and resilience, suggesting practical impact for planetary defense and space-object monitoring.

Abstract

The potential for catastrophic collision makes near-Earth asteroids (NEAs) a serious concern. Planetary defense depends on accurately classifying potentially hazardous asteroids (PHAs), however the complexity of the data hampers conventional techniques. This work offers a sophisticated method for accurately predicting hazards by combining machine learning, deep learning, explainable AI (XAI), and anomaly detection. Our approach extracts essential parameters like size, velocity, and trajectory from historical and real-time asteroid data. A hybrid algorithm improves prediction accuracy by combining several cutting-edge models. A forecasting module predicts future asteroid behavior, and Monte Carlo simulations evaluate the likelihood of collisions. Timely mitigation is made possible by a real-time alarm system that notifies worldwide monitoring stations. This technique enhances planetary defense efforts by combining real-time alarms with sophisticated predictive modeling.

A multi-model approach using XAI and anomaly detection to predict asteroid hazards

TL;DR

The paper addresses the hazard prediction of near-Earth asteroids (PHAs) by proposing a hybrid ML-DL framework that fuses textual orbital/numerical data and image data from telescope observations. It integrates Explainable AI (SHAP, LIME, Grad-CAM) and anomaly detection (Isolation Forest, One-Class SVM, Autoencoders) into a weighted ensemble of models (textual: LR, DT, RF, SVM, KNN, Naive Bayes, XGBoost; visual: CNN, EfficientNetB0, VGG16) to deliver real-time hazard predictions and alerts. Advanced preprocessing (imputation, outlier removal, scaling, PCA, RFE, augmentation) and a real-time alert system (API access, dashboards, automatic notifications) are central to robustness and timeliness. The results show top accuracies up to 99% for text-based classifications and 98–99% for image-based classifications, with XAI and anomaly detection enhancing interpretability and resilience, suggesting practical impact for planetary defense and space-object monitoring.

Abstract

The potential for catastrophic collision makes near-Earth asteroids (NEAs) a serious concern. Planetary defense depends on accurately classifying potentially hazardous asteroids (PHAs), however the complexity of the data hampers conventional techniques. This work offers a sophisticated method for accurately predicting hazards by combining machine learning, deep learning, explainable AI (XAI), and anomaly detection. Our approach extracts essential parameters like size, velocity, and trajectory from historical and real-time asteroid data. A hybrid algorithm improves prediction accuracy by combining several cutting-edge models. A forecasting module predicts future asteroid behavior, and Monte Carlo simulations evaluate the likelihood of collisions. Timely mitigation is made possible by a real-time alarm system that notifies worldwide monitoring stations. This technique enhances planetary defense efforts by combining real-time alarms with sophisticated predictive modeling.

Paper Structure

This paper contains 24 sections, 12 figures, 8 tables, 1 algorithm.

Table of Contents

  1. Introduction
  2. RELATED PRIOR WORKS
  3. Contribution of the work
  4. Problems Addressed in the Current Work
  5. Solution Proposed
  6. Novelty of the proposed Work
  7. METHODOLOGY
  8. Dataset Acquisition
  9. Advanced Data Preprocessing Techniques
  10. Textual Data Preprocessing
  11. Image Data Preprocessing
  12. Multi-Model Approach for Asteroid Classification and Hazard Prediction
  13. Rationale for Algorithm Selection in Asteroid Prediction
  14. Proposed Hybrid Algorithm for Asteroid Classification and Hazard Prediction
  15. Explainable AI (XAI) and Anomaly Detection in Asteroid Classification
  16. ...and 9 more sections

Figures (12)

  • Figure 1: Workflow diagram of the hazardous prediction with XAI and Anomaly detection
  • Figure 2: Data Distribution of Asteroid dataset
  • Figure 3: Sample image data of Asteroid
  • Figure 4: Data Preprocessing on Numerical and Image data
  • Figure 5: KDE distribution before preprocessing
  • ...and 7 more figures