Analyse this


COMMERCIAL PROFILE - UNIVERSITY COLLEGE DUBLIN:Business analytics can be applied to anything from election predictions to conflict resolution, and UCD offers the ideal Masters programme for anyone looking to break into this increasingly flexible area of science

IMAGINE A SCIENCE that could accurately predict the results of elections, assist political parties to win elections, make the trains run on time, resolve intractable and sometimes violent conflicts, identify betting scams, help football teams win matches and advise on personal injury insurance claims. This may sound like the stuff of science fiction but it is very much reality and is being taught in a new and highly successful MSc course in UCD.

The UCD MSc in Business Analytics is a one-year full-time programme or two-year part-time programme designed to give participants knowledge of and experience with current quantitative analysis techniques. The programme develops students' computer software skills to implement these techniques in a business problem-solving environment and develops an understanding of the theory and practice of allied information systems subjects.

If this sounds a little dry, it is anything but. The course, which is a year old, has already exceeded all expectations in terms of demand. The course evolved out of the Masters in Management Science programme in UCD and demand was expected to be in line with the 15 or so annual intake into the old course according to incoming business analytics course director Dr Sean McGarraghy.

"We had 43 participants on the course last year, about two-thirds of them part-time," he says. "We are expecting about the same number this year. I think this is an indication of the pent-up demand there was for a programme of this nature. We have had particularly strong interest from industry with a lot of companies facilitating their staff in taking the part-time option."

Given the clear benefits of business analytics, particularly in a recession, this strong interest is not surprising. "A recent study by Accenture found that 57 per cent of companies surveyed had no analytical tools available to them but that three-quarters of those companies intended to invest in them. And this was three months into a recession," McGarraghy says.

A further study carried out by IBM in April of this year found that half of all business leaders surveyed did not have access to the information required for them to do their job and were looking for tools to help them with this. Furthermore, eight out of 10 business leaders said they were making key decisions with missing or untested information.

"It is for these reasons that businesses have been coming to the universities over the past five years saying they needed a science to help them with decisions," McGarraghy says. "What they wanted was a science that would help them with practical decision making, and that's what we are offering in the MSc in Business Analytics programme."

Business analytics is by no means new nor is it strictly confined to business. At heart it involves the mathematical modelling of the various different factors which can influence the outcome of an event or set of events. Using these models it is then possible to put a series of "if scenarios" and see what happens if various different decisions are made.

The predictive power of the science should not be underestimated. For example, Allan Lichtman, a quantitative historian and professor at the American University in Washington DC, predicted the outcome of last year's US presidential election two years before it happened. He did this through the development of what he calls the 13 keys to presidential elections; the fundamental factors that influence the outcome.

The party or candidate who turns enough keys their way will win and open the door to the White House. And he's never been known to be wrong. Even in 2000 he predicted accurately that Al Gore would win the greater share of the popular vote, although George Bush won on the US electoral college system. Indeed, Lichtman has correctly predicted the outcome of every US presidential election since 1984, including George Bush Snr's dramatic comeback win in 1988 and the Clinton win against the same candidate four years later.

His analysis is based on a statistical pattern-recognition algorithm for predicting earthquakes developed by Russian seismologist Vladimir Keilis-Borok.

Another practical example of business analytics in use was a recent project implemented by Netherlands Railways who were finding themselves overwhelmed by growing passenger numbers. The Netherlands is unusually reliant on its rail system to bring its commuters to and from work. More than half of the country's 16 million citizens use the train as frequently as daily to get to and from work, while more than 70 per cent of people who commute from one city to another do so by train.

Unfortunately, the system became a victim of its own success and demand began to overwhelm available capacity, causing delays, cancellations and significant economic and social problems in the process.

Netherlands Railways turned to analytics to address the issue and came up with a plan to increase the number and configuration of trains, rebuild important sections of track, reschedule crews and, critically, revamp the timetable. The results were a dramatically improved service within a short period of time followed by international acclaim and awards for the rail company. Iarnród Éireann and Dublin Bus take note.

Another application of the science is in conflict resolution. Instead of using a human mediator or third party to assist in the resolution of the issues, you use a computer. The computer analyses the starting positions and bottom lines of both sides and either says no deal is possible or comes up with a series of workable compromises, with the parties to the conflict knowing they are dealing with a tireless mediator who cannot be swayed by emotion or other factors.

One of the current ongoing projects at UCD is an analysis into what constitutes "suspicious betting patterns". The aim is to develop a tool that would alert bookmakers to the probability that some form of scam is afoot. Crooked syndicates try their best to make their betting patterns look as normal as possible. "It's an arms race," says McGarraghy. "Each side is trying to be smarter than the other all the time. The work being done by one of our students at present could help tip the balance in favour of the bookies."

Another project is looking at personal injuries claims and the risks associated with contesting them.

"There is always the risk that if you go to court you might lose," says McGarraghy. "However, sometimes there are cases which can lead to what we call phase-changing events. In this instance these would be court judgements that set the bar much higher for future damages. Using business analytics, insurance companies can get better guidance on their decision making when it comes to which cases to fight and which ones to settle."

But he cautions that people should be aware of the science's limitations. "It's the old story of garbage in, garbage out," he says. "You have to look at it critically and understand its limitations."

The MSc in Business Analytics is suitable for graduates from a wide variety of disciplines with an aptitude for mathematics or computerised problem solving. The programme can lead to careers in software design, management consultancy, systems analysis, financial analysis, operations research, information technology and logistics management. The programme is part funded by the National Development Plan which covers much of the fees.

For further information see smurfitschool/businessanalytics or call 01 7168934