For UWM updates on the COVID-19 pandemic, visit uwm.edu/coronavirus.

Data Analytics

Big Data, Data Science, and Data Analytics are all ways of referring to the value of extracting useful information from the plentiful datasets now available across a wide range of fields and areas of natural and human activity. The ability to leverage data to improve understanding has always been important, but is becoming increasingly so as data becomes more readily available. The application of data analytics to a problem usually unfolds in the following sequence:

  1. What is the question/problem?
  2. What data are available that can be used to address the question/problem?
  3. What techniques (existing or entirely new) can be employed, using these data, to address the question/problem?
  4. How robust is the result? What is needed to improve this?

An example of the use of data analytics, showing the dependence of peak daily electricity demand on temperature and humidity (heat index, HI), time of day (hr) and cloud cover (fractional coverage from 0 to 1).

Most data scientists require several skills: disciplinary expertise, higher level mathematics and statistics, and computer programming. At UWM, led by Prof. Paul Roebber, we are engaged in applying the principles of data science to atmospheric science questions including the prediction of peak electricity demand (see included figure); the generation of improved deterministic and probabilistic nowcasts and short-range forecasts of precipitation, convection, severe weather, and tropical storm intensity; week 3-4 temperature forecasts; as well as questions in other fields such as professional sports (e.g., “Is there such a thing as momentum in the NFL or the NBA?”; “What is the contribution of penalty calls to home field advantage in the NFL?”). Contact us if you would like to learn more.