Forecasting and regression are two of the most widely used techniques in business analytics. Excel and JMP provide accessible platforms for modeling, allowing students to fit, test, and interpret models without requiring advanced coding skills. With AI as a companion, analysts can move beyond technical output and translate diagnostics into plain English explanations, making insights actionable for managers and executives.

Objectives

By the end of this module, students will be able to:

• Fit and interpret a simple regression model or time-series forecast.
• Use AI to explain diagnostics such as R², residual plots, and overfitting risks in clear, business-friendly terms.
• Communicate both the limits of models and practical recommendations for decision-makers.

Lecture & Discussion

This session explores the difference between forecasting (predicting future demand or trends) and classification (sorting data into categories). Students will examine the concepts of validation and data leakage, ensuring models are tested on unseen data to avoid overly optimistic performance. The discussion also introduces uplift modeling, which focuses on measuring incremental impact rather than overall correlation, a key concept for targeted marketing and resource allocation.

Hands-On Exercise

Students will conduct a demand forecasting exercise in Excel/JMP, using regression or time-series tools to predict sales. They will learn to generate diagnostic plots, assess residual patterns, and evaluate assumptions. Using Elevate AI, they will draft an executive readout—a concise summary of the forecast with key risks—and a risk register identifying uncertainties such as market shocks, supply delays, or seasonality. This exercise emphasizes not just building a model, but also communicating assumptions, limitations, and actionable “what-if” scenarios.

Lab 6: Forecast Brief

For Lab 6, students will produce a “Forecast Brief”. This deliverable includes:

• Key plots (trend line, forecast horizon, residuals).
• A model snapshot (coefficients, R², assumptions).
• A “what-if” table exploring scenarios such as price changes or demand shifts.
• A short narrative summarizing the model’s business relevance.

Midterm Practical (End of Week 6)

The midterm serves as an individual mini-case that integrates the skills learned in the first half of the course. Students will be given a dataset and tasked with:

• Writing SQL queries to prepare and explore the data.
• Building a Tableau dashboard to visualize key KPIs and insights.
• Interpreting outputs with AI assistance permitted only for explanation (not for code generation), with required disclosure of usage.

Lesson Summary

Forecasting and regression are fundamental techniques in business analytics, accessible through platforms like Excel and JMP, which enable students to model and interpret data without advanced coding skills. By integrating AI, analysts can provide clear, business-centric explanations of diagnostic results, making insights actionable for decision-makers.

Objectives for students by the end of the module include:

• Fitting and interpreting simple regression models or time-series forecasts.
• Using AI to explain diagnostics like R², residual plots, and overfitting risks in business-friendly terms.
• Communicating model limitations and practical recommendations effectively.

The lecture and discussion cover topics such as forecasting (predicting future trends) versus classification, validation, data leakage, and uplift modeling for measuring incremental impact. Hands-on exercises involve demand forecasting in Excel/JMP, examining diagnostic plots, and creating executive summaries with Elevate AI.

In Lab 6, students will create a "Forecast Brief" consisting of key plots, a model snapshot, a "what-if" analysis, and a business relevance summary. The midterm practical integrates various skills learned, including SQL queries, Tableau dashboard creation, and interpreting outputs with AI assistance limited to explanation.