Skip to content

kabirjolly6/Exoplanet_Detection

Repository files navigation

🌌 Exoplanet Detection using Machine Learning

An end-to-end machine learning system for exoplanet detection using NASA astronomical datasets, designed with a strong emphasis on scientific validity, recall-oriented evaluation, and robust generalization.

The project leverages LightGBM along with a custom refit strategy (FitSearchCV) to avoid overfitting and select stable models suitable for real-world scientific discovery.


📌 Problem Statement

Astronomical missions generate massive volumes of telescope data, making manual exoplanet detection slow and impractical at scale.

Objective:
Develop an automated and reliable ML pipeline to classify celestial objects into:

  • Confirmed Exoplanet
  • Exoplanet Candidate
  • False Positive

with special focus on minimizing missed exoplanet candidates.


🚀 Key Contributions

  • 🔬 Recall-first evaluation strategy for candidate discovery
  • 🧠 LightGBM-based classification for structured astronomical data
  • 🛡️ Explicit handling of data leakage and feature integrity
  • ⚙️ FitSearchCV: custom hyperparameter refit strategy
  • 🌐 Deployed web application for prediction and retraining
  • 📊 Careful dataset selection based on label stability

📂 Datasets

🔹 KOI (Kepler Object of Interest) — Primary Dataset

  • Clean and well-labeled
  • Stable target definitions
  • Used for training and testing

🔹 K2 Dataset

  • High reported accuracy
  • Susceptible to data leakage
  • Used only for supplementary validation

🔹 TESS (TOI) Dataset

  • Large and evolving dataset
  • Labels under continuous revision
  • Not used for supervised training

🧪 Feature Engineering

  • Final feature set: 17 features
  • Preprocessing steps:
    • Removed empty and redundant columns
    • Eliminated single-value categorical features
    • Excluded target-leaking flags
  • False-positive indicators deliberately excluded to preserve model integrity

🤖 Model Details

  • Algorithm: LightGBM (Gradient Boosted Decision Trees)
  • Why LightGBM?
    • Efficient on large tabular datasets
    • Handles non-linear feature interactions
    • High recall for exoplanet candidates
  • Primary Metric: Recall (Candidate class)

🧠 Custom Refit Strategy: FitSearchCV

Instead of refitting the model with the highest validation score, we optimize for stability using:

min ( |train_score - test_score| + (1 - test_score) ) / 2

Benefits:

  • Reduces overfitting
  • Avoids underfitting
  • Encourages generalizable models
  • Better suited for scientific discovery tasks

🌐 Application Features

  • 🔍 Real-time single prediction
  • 📁 Batch prediction with downloadable results
  • 🔄 Online retraining with new datasets
  • 🧪 Interactive web interface

🔮 Future Work

  • Automated periodic retraining
  • Intelligent false-positive flagging
  • Dataset validation pipeline
  • Continuous monitoring for model drift

About

Exoplanet detection using Machine Learning Model deployed a website.

Resources

Stars

0 stars

Watchers

0 watching

Forks

Releases

No releases published

Packages

 
 
 

Contributors