Software & Tools

Technologies and frameworks powering the AstroCluster exoplanet detection system

Dataset Overview

9,564

Total Records

6,632

Training Set

1,171

Test Set

Features Used

Data source: Kaggle Exoplanets Dataset (arashnic/exoplanets) • Binary classification task predicting exoplanet candidates

Model Performance Metrics

Random Forest

AUC Score

98.83%

Recall

94.67%

AUC Score98.83%

Recall94.67%

Parameters: n_estimators=150, max_depth=5, max_features=0.8

XGBoost

AUC Score

99.01%

Recall

95.92%

AUC Score99.01%

Recall95.92%

Parameters: max_depth=5, learning_rate=0.01, subsample=0.7

Logistic Regression

AUC Score

99.40%

Recall

95.93%

AUC Score99.40%

Recall95.93%

Parameters: C=1, max_iter=400, class_weight=balanced

Cross-Validation: All models trained using Stratified K-Fold cross-validation (3-5 folds) to ensure robust performance estimates and prevent overfitting.

XGBoost

ML Library

Gradient boosting framework for classification

LightGBM

ML Library

Fast gradient boosting on decision trees

scikit-learn

ML Library

Machine learning library with various algorithms

Pandas

Data Processing

Data manipulation and analysis

NumPy

Data Processing

Numerical computing with arrays

Matplotlib

Visualization

Data visualization library

Seaborn

Visualization

Statistical data visualization

System Architecture

Data Layer

Kaggle exoplanet dataset processing with Pandas, NumPy, and feature engineering pipelines

ML Models

Ensemble approach using Random Forest, XGBoost, and Logistic Regression with cross-validation

Web Interface

Next.js 15 with React 19, TypeScript, and Tailwind CSS for responsive, modern UI

Built with modern technologies for NASA Space Apps Challenge 2025

Software & Tools

Technologies and frameworks powering the AstroCluster exoplanet detection system

Dataset Overview

9,564

Total Records

6,632

Training Set

1,171

Test Set

Features Used

Data source: Kaggle Exoplanets Dataset (arashnic/exoplanets) • Binary classification task predicting exoplanet candidates

Model Performance Metrics

Random Forest

AUC Score

98.83%

Recall

94.67%

AUC Score98.83%

Recall94.67%

Parameters: n_estimators=150, max_depth=5, max_features=0.8

XGBoost

AUC Score

99.01%

Recall

95.92%

AUC Score99.01%

Recall95.92%

Parameters: max_depth=5, learning_rate=0.01, subsample=0.7

Logistic Regression

AUC Score

99.40%

Recall

95.93%

AUC Score99.40%

Recall95.93%

Parameters: C=1, max_iter=400, class_weight=balanced

Cross-Validation: All models trained using Stratified K-Fold cross-validation (3-5 folds) to ensure robust performance estimates and prevent overfitting.

XGBoost

ML Library

Gradient boosting framework for classification

LightGBM

ML Library

Fast gradient boosting on decision trees

scikit-learn

ML Library

Machine learning library with various algorithms

Pandas

Data Processing

Data manipulation and analysis

NumPy

Data Processing

Numerical computing with arrays

Matplotlib

Visualization

Data visualization library

Seaborn

Visualization

Statistical data visualization

System Architecture

Data Layer

Kaggle exoplanet dataset processing with Pandas, NumPy, and feature engineering pipelines

ML Models

Ensemble approach using Random Forest, XGBoost, and Logistic Regression with cross-validation

Web Interface

Next.js 15 with React 19, TypeScript, and Tailwind CSS for responsive, modern UI

Built with modern technologies for NASA Space Apps Challenge 2025

Astrocluster

Software & Tools

Dataset Overview

Model Performance Metrics

Random Forest

XGBoost

Logistic Regression

Machine Learning & Data Science

XGBoost

LightGBM

scikit-learn

Pandas

NumPy

Matplotlib

Seaborn

Backend & API

Frontend & Web Interface

System Architecture

Data Layer

ML Models

Web Interface

Software & Tools

Dataset Overview

Model Performance Metrics

Random Forest

XGBoost

Logistic Regression

Machine Learning & Data Science

XGBoost

LightGBM

scikit-learn

Pandas

NumPy

Matplotlib

Seaborn

Backend & API

Frontend & Web Interface

System Architecture

Data Layer

ML Models

Web Interface