🚀 Overview

In this capstone project, I took on the role of a data scientist at Space Y, a fictional aerospace startup aiming to rival SpaceX. The mission: determine the cost of rocket launches by predicting whether SpaceX’s Falcon 9 first stage would successfully land and be reused. Using public data, I built dashboards and trained machine learning models to solve this real-world problem.

🧠 The Challenge

SpaceX significantly lowers launch costs by reusing the first stage of its Falcon 9 rockets. While traditional launches may cost over $165M, SpaceX advertises costs around $62M. Predicting successful landings could help reduce future launch costs for competitors like Space Y. The challenge was to predict landing outcomes using public mission data, without relying on rocket science—just data science.

🛠️ Process

Gathered launch data via SpaceX public API
Cleaned and formatted datasets using Python and Pandas
Performed exploratory data analysis (EDA) to identify key features
Created interactive visualizations using Plotly and Dash
Trained ML models (Logistic Regression, SVM, Decision Tree)
Evaluated and selected the best model based on accuracy and interpretability

📊 Exploratory Data Analysis

Below are some key insights from the EDA phase. I analyzed feature distributions, correlation heatmaps, and launch outcomes to inform the model design.

📈 Data Visualizations

📉 Model Performance

🏆 Tech Stack & Achievements

Technologies Used: Python, Pandas, Plotly, Dash, Scikit-learn, Jupyter Notebooks
Built a fully interactive dashboard to visualize launch metrics and trends
Performed feature engineering to improve model accuracy
Compared multiple classification models and selected Decision Tree as the most effective
Achieved accurate predictions on first stage landings using public SpaceX data
Successfully simulated real-world data science workflow from collection to model deployment

🚀 Final Thoughts

This project showcases how data science and machine learning can be used to solve complex engineering problems—without being a rocket scientist. The insights generated provide strategic value for any company aiming to optimize launch costs and compete in the new space race.

📝 Key Learning

Through this project, I gained valuable experience in data preprocessing, exploratory data analysis (EDA), and the application of machine learning algorithms to real-world problems. I learned how to handle large datasets, evaluate model performance, and create interactive dashboards for data visualization. This project reinforced my understanding of the importance of data quality, feature selection, and the iterative nature of data science workflows.

🔮 Future Improvements

In the future, I plan to enhance this dashboard with real-time data streaming to predict launch outcomes live. Additionally, I would like to integrate more advanced machine learning models, such as deep learning approaches, to improve prediction accuracy. Expanding the dashboard to include more metrics and providing a more interactive user experience would also be valuable next steps.

🔗 View Full Project

Check out the full project here.

Validation Metrics

Detailed Intelligence