SpaceX Falcon 9 Launch Analysis 🚀

IBM Data Science Capstone Project

This project analyzes SpaceX Falcon 9 launch data to study first-stage landing success and build predictive machine learning models.
It demonstrates an end-to-end data science workflow, from data collection to model evaluation.

---

📌 Project Overview

SpaceX’s Falcon 9 rocket is designed for reusability, significantly reducing launch costs.
Predicting whether the first stage will successfully land is crucial for cost estimation and mission planning.

Objective:

• Analyze historical Falcon 9 launch data
• Identify factors affecting landing success
• Build machine learning models to predict landing outcomes

---

🛠️ Tools & Technologies Used

• Python
• Pandas, NumPy
• Matplotlib, Seaborn
• SQL (SQLite)
• Folium (Maps)
• Dash (Interactive Dashboard)
• Scikit-learn
• Jupyter Notebook

---

📂 Project Structure

spaceX-Falcon-9-Launch-Analysis/ │ ├── notebooks/ │ ├── spacex-data-collection.ipynb │ ├── spacex-data-collection-api.ipynb │ ├── spacex-data-wrangling.ipynb │ ├── eda-sql-coursera_sqllite.ipynb │ ├── SpaceX_Machine Learning Prediction.ipynb │ ├── presentation/ │ └── Spacex Falcon 9 launch analysis.pdf │ └── README.md

---

🔍 Project Workflow

1️⃣ Data Collection

• Collected launch data using the SpaceX REST API
• Scraped additional information from Wikipedia
• Extracted launch site, payload, orbit, and landing outcomes

2️⃣ Data Wrangling

• Handled missing and inconsistent values
• Selected relevant features
• Created a binary target variable for landing success

3️⃣ Exploratory Data Analysis (EDA)

• SQL queries to analyze launch sites and mission outcomes
• Visual analysis using scatter plots, bar charts, and trend analysis

4️⃣ Interactive Visual Analytics

• Folium maps to visualize launch locations
• Dash dashboard for interactive exploration of payload vs success

5️⃣ Predictive Modeling

• Models used:
  • Logistic Regression
  • Support Vector Machine (SVM)
  • Decision Tree
• Hyperparameter tuning using GridSearchCV
• Best model achieved ~83% accuracy

---

📊 Results & Insights

• Payload mass and orbit significantly affect landing success
• Certain launch sites show higher success rates
• Booster reusability improves with experience over time

---

📝 Note on Empty Cells

Some notebooks contain empty cells.
These cells were intentionally left blank as part of the original IBM Data Science Capstone lab structure for optional practice and exploration.

✔️ All required tasks, analyses, visualizations, and models have been fully completed.
✔️ No mandatory content is missing.

---

📈 Future Work

• Use larger and more recent datasets
• Apply advanced machine learning models
• Deploy the trained model as a web application

---

👤 Author

Hruthikamamidala
IBM Data Science Capstone Project

---

⭐ If you found this project useful, feel free to star the repository!