This five-week, project-driven course builds production-ready data science skills end-to-end — from gathering requirements, statistical analysis, and preprocessing through building predictive and forecasting models to deploying a live application. Students work with structured, tabular data of their own choosing, applying skills to a real research or business problem. The course also examines AI industry trends and the evolving role of data science across careers and society. All work is in Python; all assignments are project-based.

Python Ecosystem and Tools

• Data & EDA: Pandas, NumPy, data-profiling, missingno
• Visualization: Matplotlib, Seaborn, Plotly
• Statistics & Modeling: SciPy, Statsmodels, Scikit-Learn, imbalanced-learn
• Forecasting: Statsmodels (ARIMA/SARIMA), Prophet
• Deployment & Collab: Streamlit, joblib / pickle, Git / GitHub, Jupyter Notebook

Disclaimer: The tools, technologies, and learning outcomes listed in this syllabus are subject to change based on industry developments, software availability, or course improvements. Students will be notified in advance of any adjustments.

Learning Outcomes

By the end of this course, students and professionals will be able to:

1. Managing the Full Data Science Project
   • Understand how a real data science project is structured — from identifying a business or research problem all the way to delivering a working solution others can use.
   • Organize and present your work clearly in Jupyter notebooks and a live, shareable web application built with Streamlit.
2. Thinking Statistically
   • Use statistics the way data scientists practice — to describe data, test ideas, and draw reliable conclusions rather than just guess.
   • Run and interpret common tests such as t-tests, chi-square, and ANOVA, and understand what p-values and confidence intervals actually mean.
   • Recognize when a regression model is trustworthy and when its assumptions are being violated.
3. Exploring & Preparing Data
   • Load datasets from files, databases, and APIs, and quickly understand what you are working with using profiling and visualization tools.
   • Clean messy data, fill gaps, remove noise, and transform variables so they are ready for analysis and modeling.
   • Build reusable, professional-grade data preparation workflows that others can reproduce and extend.
4. Building Predictive Models
   This course focuses on understanding and applying models correctly, not on deep algorithm theory.
   • Build Linear Regression models to predict numerical outcomes and understand what is driving those predictions.
   • Build Logistic Regression models to classify data and evaluate how well the model actually performs in practice.
   • Select the best model version using cross-validation and avoid the common mistake of overfitting to training data.
5. Forecasting with Time Series
   • Recognize patterns in time-based data — trends, seasons, and cycles — and prepare the data correctly before modeling.
   • Build ARIMA, SARIMA, and Prophet forecasting models and measure how accurate their predictions are.
   • Use automated tools to speed up model selection without sacrificing quality.
6. Visualizing Results & Deploying Solutions
   • Create clear, compelling charts and dashboards that communicate findings to both technical and non-technical audiences.
   • Save trained models and build interactive web apps that let others explore data and get predictions in real time.
   • Manage project code professionally using Git and GitHub, so your work is organized, version-controlled, and shareable.