SpaceX stands as a game-changer in the realm of space exploration, redefining the economics of space missions. Their groundbreaking approach to cost-efficiency is primarily showcased through their Falcon 9 rocket launches, which are priced at a competitive 62 million dollars, a stark contrast to the hefty 165 million dollar price tags of other providers. The cornerstone of this cost reduction is SpaceX's ingenious strategy of recycling the first stage of their rockets.
By successfully landing and reusing this stage for subsequent missions, they've set the stage for even more cost savings in the future. From the perspective of a data scientist working for a startup in competition with SpaceX, the main objective of our project is to devise a machine-learning model that predicts the landing outcomes of these rockets' first stages. Such a project is pivotal, as it will equip us with insights to strategically price our own launches in a bid to challenge SpaceX's market dominance. Key challenges encompass:
• Pinpointing the determinants that sway the landing results. • Understanding the intricate interplay between variables and their impact on the landing. • Deciphering the optimal conditions that bolster the chances of a successful touchdown.
- Data Collection Methodology
- SpaceX API and Web Scraping From Wikipedia
- Perform Data Wrangling
- One Hot Encoding data from Machine Learning
- Perform exploratory data analysis (EDA) using visualization and SQL
- For data visualization, I utilized various plotting libraries.
- Perform interactive visual analytics using Folium and Plotly Dash
- Perform predictive analysis using classification models
- I utilized Logistic Regression, KNN, SVC, and DecisionTreeClassifier for the model-building process in my analysis.
- Data Collection with API
- Data Collection with Web Scraping
- Data Wrangling
- EDA with SQL
- EDA with Visualization Lab
- Interactive Visual Analytics with Folium
- Predictive Analysis - Machine Learning Lab
- Final Presentation (Slide)