Forecasting and Regression

Lesson 5 – Introduction to Forecasting and Regression

This lesson introduces forecasting.  It starts by presenting qualitative, time series, and causal models.  It then explores the common qualitative forecasting approaches of the Delphi Method, Jury of Executive Decision, Sales Force Composite, and Consumer Market Survey.

Lesson 6 – Time Series Models

This lesson introduces time series data.  We then cover several quantitative time series forecasting methods presenting moving average (MA), weighted moving average (WMA) and exponential models.  As we present each type of model we show how to develop the model in Excel (Google Forms).

Lesson 7 – Metrics to Compare Forecasts

Forecast accuracy is the difference between the actual data and the forecast value. When making a decision we want the most accurate forecasting model possible, but how we measure forecast accuracy depends on how the model will be used.  We present three methods of measuring forecast accuracy, mean absolute deviation (MAD), mean squared error (MSE), and mean absolute percent error (MAPE); and discuss when each measure is most appropriate for selecting your forecast model.  Interwoven through the lesson is the development of all three metrics in Excel.  At the end, we practice model selection using our forecast models developed in Lesson 6.  We also introduce Excel’s Solver for selecting optimal weights when developing weighted moving average forecast and exponential smoothing models to minimize an error metric.

Lesson 8 – Linear Regression & Causal Models

This lesson introduces linear regression.  We start by explaining the relationship between the dependent and independent variables explored in casual models.  Next we develop a one variable regression and then expand to multiple linear regression.  We then examine regression model outputs covering: multiple R, R2, adjusted R2, F-test, coefficients, T-test, and confidence intervals.  The lesson concludes by using a regression model to make forecasts.

Lesson 9 – Seasonality and Indices

A studied event may be affected by a recurring cycle.  For example, ice cream sales increase in the summer months while decreasing in the winter.  This phenomenon is known as seasonality.  One method to adjust for seasonality is indices.  Indices have broad uses, from adjusting travel pay between cities to identifying next season’s inter collegiate athletes at risk of failing.

Homework & Section Materials

Excel Lesson 6 & 7 – Time Series Models and Metrics (download or copy to edit)

Homework 4 – Time Series Models and Metrics

Excel Lesson 8 – Linear Regression and Causal Models (download or copy to edit)

Homework 5 – Linear Regression and Causal Models

Excel Lesson 9 – Seasonality and Indices (download or copy to edit)

Homework 6 – Seasonality and Indices

1. Fee says: