IFI 8410: Introduction to Programming and Predictive Analytics for Business
Listen to the introduction
Transcript
| Speaker | Text |
|---|---|
| Alex | This is the brief on MSA eighty-fifty scalable data analysis. You’ve got the analytics basics down, but what happens when you’re suddenly staring at an absolute mountain of data? Think of this course as taking your existing skills and hooking them up to an industrial sized engine to tackle scalable analytics, which is honestly crucial for modern retail and massive industries. First, we dive into the fusion of algorithmic theory and massive scale. We’ll cover core concepts like machine learning and forecasting, but we really push you to ask, does a standard algorithm still work when the data is literally too big for one computer? No way. That’s exactly where distributed systems like Apache Spark step in to save the day. Second, get ready for hands-on, end to end retail analytics. You’re gonna build a complete solution from scratch, designing data extraction pipelines, running a baseline, and building a far superior model. And why retail? Because it generates an absurd amount of messy, complex data, making it the perfect proving ground for testing scalable systems in the wild. Finally, you’ll turn these skills into a polished professional portfolio piece. You’ll use real industry tools like GitLab for version control and the university’s ARC Supercomputing cluster. Think of ARC and GitLab as a flight simulator for data scientists. They give you undeniable proof that you can handle real-world pressures before even hitting the job market. Ultimately, MSA 8050 takes your foundational analytics knowledge and scales it up to solve massive industry problems, taking you from raw data pipelines to a professional grade presentation. |
This course introduces students to the science of business analytics and covers foundational material needed to use, apply and develop analytic solutions to real-world business challenges. Students will learn to identify the appropriate methods to collect, analyze, and visualize data and utilize data in decision making. Examples will illustrate the use of models to solve business problems such as reducing customer churn, customer segmentation, predicting market demand, forecasting stock prices, etc. Both structured and unstructured data will be used throughout the course. The course prepares students for IFI 8420, which focuses on the implications of machine learning and artificial intelligence for business strategy and creating business value.
Schedule
| Session | Date | Topic | In Class | Project Milestone |
|---|---|---|---|---|
| 1 | 2026-08-26 | Course Overview & Scale | ACT01 | Optional background & project preference survey (ungraded) |
| 2 | 2026-09-02 | Problem Formulation & Planning | ACT02 | M01: Team roster, project choice, and 1-page execution plan |
| 3 | 2026-09-09 | Recommenders & Frequent Patterns | ACT03 | |
| 4 | 2026-09-16 | Graph Basics & Connectivity | ACT04 | M02: Data understanding & initial ETL report (schema, data quality, first ETL run) |
| 5 | 2026-09-23 | Time Series & Many-Series Forecasting | ACT05 | |
| 6 | 2026-09-30 | Factorization & Sequential Patterns | ACT06 | M03: Baseline system implemented and evaluated (ETL+features, metrics, runtime notes) |
| 7 | 2026-10-07 | Communities, Triangles & Structure | ACT07 | |
| 8 | 2026-10-14 | Parallel ETL & Distributed Patterns | ACT08 | |
| 9 | 2026-10-21 | Model Selection & Large-Scale Experiments | ACT09 | M04: Improved system implemented, with comparative technical and quality metrics and initial trade-off analysis |
| 10 | 2026-10-28 | Optimization & Iterative Algorithms | ACT10 | |
| 11 | 2026-11-04 | Robustness, Drift & Monitoring | ACT11 | |
| 12 | 2026-11-11 | Algorithm-System Trade-offs & Cases | ACT12 | M05: Draft final report and slides with full baseline vs improved metrics and narrative |
| 13 | 2026-11-18 | Synthesis & Presentation Prep | ACT13 | |
| 14 | 2026-12-02 | Last day of class: Presentations | ||
| 2026-12-09 | -- no class -- | M06: Final reports and video |
Documents
Topics
- Course Overview & Scale
- Problem Formulation & Planning
- Recommenders & Frequent Patterns
- Graph Basics & Connectivity
- Time Series & Many-Series Forecasting
- Factorization & Sequential Patterns
- Communities, Triangles & Structure
- Parallel ETL & Distributed Patterns
- Model Selection & Large-Scale Experiments
- Optimization & Iterative Algorithms
- Robustness, Drift & Monitoring
- Algorithm-System Trade-offs & Cases
- Synthesis & Presentation Prep
- Last day of class: Presentations
Project
The course projects comprise a curated set of retail analytics problems designed with consistent scope, data scale, and technical rigor, forming the primary analytical work of the course. Each project follows a structured sequence of milestones that guide the development of an end‑to‑end scalable data analytics system, beginning with data understanding and ETL/ELT pipeline construction and progressing through baseline and improved modeling, systematic evaluation of analytical quality and system performance, and comprehensive reporting. Projects are developed and versioned using the internal GitLab repository and executed on the ARC to ensure reproducibility in a realistic computing environment. The milestone sequence culminates in a final report and presentation, producing a polished technical artifact suitable for inclusion in a professional portfolio.
Project Milestones
| Milestone | Due |
|---|---|
| M01: Team roster, project choice, and 1-page execution plan | 2026-09-02 |
| M02: Data understanding & initial ETL report (schema, data quality, first ETL run) | 2026-09-16 |
| M03: Baseline system implemented and evaluated (ETL+features, metrics, runtime notes) | 2026-09-30 |
| M04: Improved system implemented, with comparative technical and quality metrics and initial trade-off analysis | 2026-10-21 |
| M05: Draft final report and slides with full baseline vs improved metrics and narrative | 2026-11-11 |
| M06: Final reports and video | 2026-12-09 |