Kepler Objects of Interest (KOI) Analysis: Classification Project

Description

This project analyzes data on Kepler Objects of Interest (KOIs) from the NASA Exoplanet Archive. The primary goal is to develop a statistical model to classify KOIs as either potential exoplanet CANDIDATEs or FALSE POSITIVEs based on their observed properties. The analysis utilizes the Q1-Q17 Data Release 25 (DR25) KOI table, chosen for its homogeneity and automated vetting process suitable for statistical analysis.

The project involves several stages:

1. Data loading and preparation.
2. Exploratory Data Analysis (EDA) through visualization.
3. Dimensionality reduction using Principal Component Analysis (PCA).
4. Predictive modeling using Generalized Linear Models (Logistic Regression).

Data

The primary dataset used is:

• data/q1_q17_dr25_koi.csv: Contains the Kepler Objects of Interest cumulative table from Data Release 25 (Q1-Q17). This dataset was downloaded from the NASA Exoplanet Archive.

Project Structure

progetto_SIL/ ├── data/ │ ├── q1_q17_dr25_koi.csv # Primary DR25 KOI dataset ├── data_preparation.Rmd # R Markdown script for data loading, cleaning, and initial summary ├── data_visualization.Rmd # R Markdown script for exploratory data analysis visualizations and PCA analysis ├── model.Rmd # R Markdown script for models testing └── model_pca.Rmd # R Markdown script for models testing using PCA