DS_SpaceX_F9

In this project, we aim to predict whether the Falcon 9 rocket's first stage will successfully land.

Business Impact: Achieving a successful landing of SpaceX's rocket first stage could dramatically reduce launch costs from $165 million to $62 million, offering a $103 million advantage over competitors.

Technical Skills:

✅ Data Collection using API (SpaceX API), and Web-scraping using BeautifulSoup

✅ EDA Using SQL, Pandas, and Matplotlib

✅ Interactive Visual Analytics and Dashboards (Folium & Plotly Dash)

✅ Predictive Analysis using four different Classification Models (Decision Tree, K-Nearest Neighbor, Support Vector Machine, and Logistic Regression

---

SpaceX offers these launches for $62 million, much cheaper than competitors' prices, which start at $165 million. This price difference is largely because SpaceX can reuse the first stage. Knowing whether the first stage will land allows us to figure out the launch cost. This information could help other companies compete with SpaceX for launching rockets.

Questions for analysis:

1. Which factors most significantly contribute to the successful landing of Falcon 9's first stage?
2. What influence do specific launch parameters (such as payload mass, orbit type, and launch site) have on the success rate of Falcon 9's first-stage landing?
3. Which Classification Model best performed for this data set?

Conclusion:

• Payload, Orbit Type and Launch Site play huge factor to the successful landing of F9’s first stage.
  • Payload Mass between 2,000 to 4,000 kgs has the highest success rate for all sites.
    • However, for Orbits Polar, LEO, and ISS → they have positive positive landing rates for heavier payloads.
  • KSC LC-39A Launch Sites has the highest success rate (77%), with 10/13 successful landings.
  • SSO Orbit had no unsuccessful landing.
• By focusing on those 3 factors, we can improve the success rate of F9 future launches.
• Decision Tree Model best performed for this dataset with 89% accuracy.