Falcon 9
The project analyzes SpaceX Falcon 9 launch data, and uses machine learning models to predict landing outcomes.
Kickstarter
This project analyzes SpaceX launch data, beginning with data collection via SpaceX API and web scraping, followed by data cleaning and preparation. It includes exploratory data analysis (EDA) using SQL and visualization techniques to identify patterns and a machine learning component for forecasting successful launches. The final stage is the creation of an interactive dashboard with Dash to display the analysis results.
The Details
Data Collection
Data Wrangling
EDA
Feature Engineer
Model Development
Dashboard Creation
I built a Dash dashboard to make the analysis accessible and interactive, allowing users to explore the data, view predictions, and gain insights through various visualizations.
Results and Insights
Flight Number vs. Launch Site
Payload vs. Launch Site
Payload vs. Launch Site
Payload vs. Launch Site
The Model
I developed machine learning models, including logistic regression, decision trees, and random forests, to predict launch success, using cross-validation and grid search for hyper-parameter tuning. The models were evaluated using accuracy, precision, recall, and F1-score, helping to select the best-performing model for deployment.
The predictive analysis aimed to predict the outcome of SpaceX launches using features like Payload Mass, Orbit, Launch Site, and Landing Outcome. After preprocessing the data and converting categorical variables using one-hot encoding, three classification models (Decision Tree, Logistic Regression, and K-Nearest Neighbors) were trained and evaluated using the F1 score. The Decision Tree Classifier, optimized with Grid Search Cross-Validation, emerged as the best-performing model with an improved F1 score of 0.92.