Applied Machine Learning for Business

Developing applied machine learning skills in business contexts, this course covers customer segmentation and lifetime value, churn prediction with temporal features, credit scoring and fraud detection under class imbalance, uplift modeling for targeting, and recommender systems. The second half expands into text analytics, transfer learning for images, sequential models, transformers, and foundation model applications including an exercise building AI agents for market research.

Practical trade-offs are central throughout: when to use complex models versus simpler approaches, how to handle imbalanced data, how to evaluate recommenders beyond accuracy, and when API-based AI makes sense versus custom solutions. Four problem sets provide hands-on experience; a group project integrates technical and business analysis.

This course develops applied ML skills in business contexts. We cover: customer segmentation and lifetime value, churn prediction with temporal features, credit scoring and fraud detection under class imbalance, uplift modeling for targeting, and recommender systems. The second half expands into text analytics, transfer learning for images, sequential models, transformers, and foundation model applications including an exercise building AI agents for market research. The curriculum emphasizes practical tradeoffs: when to use complex models versus simpler approaches, how to handle imbalanced data, how to evaluate recommenders beyond accuracy, and when API-based AI makes sense versus custom solutions. Four problem sets provide hands-on experience; a group project integrates technical and business analysis.

After completion of this course, students will be able to:

• Apply ML methods to problems framed by business decisions, costs, and constraints
• Work with diverse data types (text, images, temporal sequences, imbalanced datasets)
• Understand key conceptual frameworks: platform economics, privacy-utility tradeoffs, scaling laws for foundation models
• Make informed choices about method selection based on problem characteristics

As a result of this course, students will be better prepared to:

• Pursue capstone or thesis work in business analytics
• Communicate technical work to non-technical stakeholders
• Navigate applied ML projects across multiple data types
• Integrate ML, statistics, economics, and programming for real-world problems
• Transition into applied business roles, serving as a bootcamp for CS majors seeking industry positions in analytics, product, or data science

Info not available.

Info not available.

• Problem Sets: 20% (4 × 5%, due Weeks 4, 7, 10, 11)
• Midterm Examination: 20% (Week 7)
• Final Examination: 30% (Week 16)
• Group Project: 20% (Proposal 5%, Report 10%, Presentation 5%)
• Attendance: 10%