IBM Data Science Capstone Project

Project Overview

This repository contains my capstone project for the IBM Data Science Professional Certificate, completed in April 2024, prior to beginning my Master’s in Data Science.

The project demonstrates a complete end-to-end data science workflow, including data collection, preprocessing, exploratory analysis, and machine learning modeling.

While the dataset is relatively small, this project establishes a strong foundation and serves as a baseline for more advanced work developed during my graduate studies.

---

Objective

The goal of this project is to predict whether the first stage of a Falcon 9 rocket will land successfully, which has direct implications for launch cost estimation.

---

Key Skills Demonstrated

• Data collection (API & web scraping)
• Data wrangling and preprocessing
• Exploratory Data Analysis (EDA)
• Data visualization
• Geospatial analysis (Folium)
• Interactive dashboards (Dash)
• Machine learning modeling and evaluation

---

Project Presentation

View Project Presentation (PDF)

---

Important Note About the Notebooks

The Jupyter notebooks in this repository were provided as guided lab assignments by IBM as part of the capstone project.

• The structure and instructions were provided by IBM
• All code implementations, analysis, and answers were completed by me
• The notebooks were submitted for grading as part of the certification

The final presentation included in this repository follows the structure provided by IBM; however, all analysis, results, and content were completed independently by me.

The datasets used in this project were provided by IBM through publicly accessible cloud storage links.

---

Repository Structure

.
├── data/
│   ├── dataset_part_1.csv
│   ├── dataset_part_2.csv
│   ├── dataset_part_3.csv
│   ├── spacex_launch_dash.csv
│   └── spacex_launch_geo.csv
│
├── notebooks/
│   ├── Collecting_the_Data_API.ipynb
│   ├── Webscraping_Records_from_Wikipedia.ipynb
│   ├── Data_Wrangling.ipynb
│   ├── EDA_using_SQL.ipynb
│   ├── Data_Visualization.ipynb
│   ├── Launch_Sites_Locations_Analysis_with_Folium.ipynb
│   ├── Machine_Learning_Prediction.ipynb
│   └── Dashboard_Application_with_Dash.ipynb
│
├── reports/
│   └── IBM_Capstone_Presentation_Report.pdf
│
├── requirements.txt
└── README.md

---

Notebooks Description

• Collecting the Data (API)
  Data collection using the SpaceX API
• Web Scraping Records from Wikipedia
  Data extraction using web scraping
• Data Wrangling
  Data cleaning and preprocessing
• EDA using SQL
  Exploratory analysis using SQLite3
• Data Visualization
  Visual analysis using matplotlib, seaborn, and plotly
• Launch Sites Location Analysis using Folium
  Geospatial visualization and mapping
• Dashboard Application using Dash
  Interactive dashboard for data exploration
  The dashboard implementation is included in the Jupyter notebook
• Machine Learning Prediction
  Model training, tuning, and evaluation

---

Technologies & Libraries

Python Libraries

• pandas
• numpy
• matplotlib
• seaborn
• plotly
• sqlite3
• dash
• folium
• requests
• datetime
• os

Machine Learning (scikit-learn)

• train_test_split
• GridSearchCV
• LogisticRegression
• SVC
• DecisionTreeClassifier
• KNeighborsClassifier

---

Setup

Install dependencies:

pip install -r requirements.txt

---

Key Takeaways

• Built a complete data science pipeline from raw data to predictive modeling
• Developed interactive visualizations and dashboards for insights
• Applied geospatial analysis to understand launch site behavior
• Compared multiple machine learning models for classification performance

---

Future Improvements

• Apply ensemble methods such as Random Forest and Gradient Boosting
• Use larger and more recent datasets
• Deploy the dashboard as a live web application