Explainable Machine Learning for Exoplanet Candidate Classification
Analyze real NASA exoplanet data, predict whether a celestial object is an exoplanet candidate, and understand exactly why the AI made its decision.
A full research pipeline — from raw NASA telemetry to explainable predictions.
Explore real astronomical observations from the NASA Exoplanet Archive.
Handle missing values and prepare data automatically.
Train multiple classification algorithms and compare performance.
Understand every prediction using SHAP explanations.
Confusion matrix, ROC curve, feature importance, and EDA charts.
Eight steps from telescope data to a trusted, interpretable prediction.
Launch the interactive dashboard and run a full classification pipeline in seconds.