One of such application area is Quality Control, a subject that has seen a strong revival during the last decade. As an example in this area, Albers comments: 'One application is the use of control charts. The outcome of a large-scale production process needs to be monitored and for this purpose at regular points in time a specific characteristic of a sampled product is registered. These outcomes are plotted and if they start to fall outside certain control limits, the process is declared 'out of control'. The machine is then stopped for repair. This is fine if the process had actually gone out of control. But if it was still operating as it should be, the out-of-control signal really is 'bad luck'. Clearly, one should try to strike the proper balance between on the one hand letting out-of-control processes carry on, and on the other hand interfering with processes with which nothing is wrong. Hence setting proper control limits is a delicate matter! Fortunately, the increased ease of computation because of easily accessible computing power, together with the availability of much more powerful methods from mathematical statistics, allows considerable improvements'.
One PhD student participant in this research is Sri Nurdiati, from Indonesia. Here, she tells us why she chose to join the group: 'The project can be used directly or indirectly to improve the quality of the industrial process and products. It's a very interesting project and it's very relevant for the needs of the industrial sector in my country'.
Another important field for the group is that of Finance and Insurance. More and more complicated products are offered on both the financial and the insurance markets, with very intricate mutual relations. Again, Albers illustrates: 'Suppose an insurance company expects say 4 million guilders every year as the total claim amount in a given portfolio. But they are well aware that this amount will vary, due to stochastic fluctuation. So they might consider to buy reinsurance: if in a given year the loss exceeds say 6 million, the surplus will then be covered by a reinsurer (usually some consortium of insurance companies). Of course, the original insurance company will have to pay a premium for this coverage. But what is a fair price? Again, advanced methods from probability and mathematical statistics can help to solve these types of problems.'
The other PhD student of the group Isnani Darti, also an Indonesian, explains: 'I chose the project of Insurance Mathematics becauseI think it's very interesting and important for science as we can here develop knowledge in relation to insurance'.
Another PhD position is still open at the moment and there is a strong possibility that it will be filled by another foreigner. Professor Albers explains this as follows: 'The demand for applied mathematicians (and certainly statisticians) is considerably larger than the output the universities are offering the last few years. Students get jobs in industries like Philips and AKZO, banks and insurance companies, consultancy and market research agencies, or (semi-)government institutions like the national railway company, so they only rarely choose to do PhD work.'
