2. Predictive modeling – what will happen next?
3. Forecasting and extrapolation – what if these trends continue?
4. Statistical analysis – why is this happening?In this column I would like to concentrate on the first level. I start with mentioning an influential book on the impact of Optimization in our society: The Optimization Edge; Reinventing Decision Making to Maximize All Your Company’s Assets (2011), written by Steve Sashihara, president and CEO of Princeton Consultants Incorporated. In the first part of the book he discusses the following question: ‘Why do some companies become industry leaders, while others never rise to the top?’. ‘Why is McDonald’s much more successful than Roy Rogers?’ Same question for Walmart versus Kmart, Marriott versus Howard Johnson’s, Google versus Yahoo!, UPS versus Airborne Express, Amazon versus Borders.
The author of this book shows in a convincing way that the basic answer to this question is that the leaders of industry ‘posses an ability to make complex decisions faster, more accurately, and more consistently than their competitors’. These companies are able to do so because they are big users of Optimization!
It is also striking to see that Optimization not only plays a crucial role in the commercial world, but also in many ‘public interest’ environments. I give two examples that I know from my own applied research. The first example is the determination of the new safety standards of the dike ring areas in the Netherlands, in which Optimization played a crucial role. The second example is the crucial role of Optimization in cancer treatment. Nowadays, hospitals are using optimization techniques to find optimal radiation treatment plans for patients.
An important development that makes the use of Optimization more crucial is the ICT development. In the past the lack of data was a big problem for many companies, but nowadays often huge amounts of data are available. A few weeks ago I visited the IBM research lab in Dublin, and they showed me many impressive examples in this respect. One example is that they use the huge amount of data of mobile telephones in a city to make transport ‘smarter’. Notice that this ‘Big Data’ development also leads to new optimization challenges. Solving really large-scale optimization problems is one of them. In most applications the speed is important. It is more important to obtain a robust, good solution in time, than an exact optimal solution too late.
My advice to the econometrics students is to read the above mentioned book by Sashihara. I also would like to point out that Steve Shashihara is one of the keynote speakers of the OR2013 Conference, which will be held from Tuesday 3 through Friday 6 September 2013 in the city of Rotterdam (for more information see http://www.or2013.org).
For the econometrics students it is good to know that the future of Optimization, and Operations Research in general, is bright! Even in these challenging economic times, there is still much demand for students who are graduated in Optimization/Operations Research. The expectation is that this will even increase in the future!
Tekst by: Dick den Hertog