This project is the grand finale of the IBM Data Science Professional Certificate. It brings together everything learned across the specialization into one real-world, end-to-end data science project.
SpaceX has revolutionized space travel by drastically cutting launch costs through rocket reusability. A Falcon 9 launch costs about $62 million, compared to over $165 million from competitors. The secret? Successfully landing and reusing the first stage booster.
Our mission: 🔭 Use data science and machine learning to predict whether the first stage of a Falcon 9 will successfully land. This prediction directly translates into understanding the economics of a SpaceX launch.
- How do factors like payload mass, launch site, flight history, and orbit type influence landing success?
- Have landing success rates improved over time?
- Which machine learning algorithm performs best for this binary classification problem?
- Pulled structured launch data using the SpaceX REST API
- Collected additional details via Wikipedia web scraping
- Cleaned and filtered raw data
- Handled missing values
- Applied One Hot Encoding to prepare features for machine learning
- Conducted visual EDA with Matplotlib & Seaborn
- Executed SQL queries for structured insights
- Built geo-visualizations with Folium
- Created interactive dashboards using Plotly Dash
- Trained and fine-tuned multiple classification models
- Compared their performance to select the most accurate predictor
- A data-driven understanding of what drives successful Falcon 9 landings
- Interactive dashboards for visual exploration
- A machine learning model that can predict landing success with high accuracy
🔥 This project isn’t just about rockets—it’s about bringing together data wrangling, EDA, visualization, and machine learning into a single, cohesive data science solution.