I noticed that some economic papers were published with multivariate regression models that had a low r-squared but a significant p-value for the predictor variables. I thought this odd, so I found some discussion on the web. I then played around with a simple linear regression model to see how such a situation might arise.

Fit a regression model to sales. Choose which predictors to include.

Tsing Hua Business Analytics and Forecasting Unit 5.7

Smoothing methods for series with multiple seasonal cycles

Trying to get hourly data to work based on Shmueli book: Practical Time Series. Example is on p. 105, but while it works for daily data, it does not generalize to hourly data. Why?

BUSINESS ANALYTICS USING FORECASTING NATIONAL TSING HUA UNIVERSITY

For BUSINESS ANALYTICS USING FORECASTING NATIONAL TSING HUA UNIVERSITY

For: BUSINESS ANALYTICS USING FORECASTING NATIONAL TSING HUA UNIVERSITY

Assignment 3.9 for Business Analytics using Forecasting at National Tsing Hua University

For FutureLearn Business Analytics using Forecasting Project. National Tsing Hua University

This is a “back of the envelope” Extreme Value Analysis utilizing weather event data obtained from the U.S. National Oceanic and Atmospheric Administration’s (NOAA) storm database for the years from 1950 to 2011. The weather event categories utilized are sourced from the National Weather Service (NWS). Using the storm database and Extreme Value Analysis techniques, I address questions about what extreme weather events imply for the United States in terms of catastrophic mortal harm as well as catastrophic injury and property damage.