Göttingen, University, Diss., 2002.

University, Diss., 2000--Tübingen.

University, Diss., 2000--Dortmund.

Universiẗat, Diss., 2003--Bremen.

Ulm, Univ., Diss., 2009.

We consider autoregressive neural network (ARNN) processes driven by additive noise. Sufficient conditions on the network weights (parameters) are derived for the ergodicity and stationarity of the process. It is shown that essentially the linear part of the ARNN process determines whether the overall process is stationary. A generalization to the case of integrated ARNN processes is given. Least squares training (estimation) of the stationary models and testing for non-stationarity are discussed. The estimators are shown to be consistent and expressions on the limiting distributions are given. / Series: Working Papers SFB "Adaptive Information Systems and Modelling in Economics and Management Science"

Despite the lack of a precise definition of volatility in finance, the estimation of volatility and its prediction is an important problem. In this paper we compare the performance of standard volatility models and the performance of a class of neural models, i.e. mixture density networks (MDNs). First experimental results indicate the importance of long-term memory of the models as well as the benefit of using non-gaussian probability densities for practical applications. (author's abstract) / Series: Report Series SFB "Adaptive Information Systems and Modelling in Economics and Management Science"

Karlsruhe, Universität, Thesis (doctoral), 2000.

We study the product development process in an artificial firm using two different incentive schemes (market share/production costs vs. life cycle return). In the product development process, we compare a trial and error search to the House of Quality approach. In our study, we focus on tactical decision making within a stable environment, given resources (production function) and knowledge base. The knowledge base is represented by neural networks which are trained on the basis of prototype data. This knowledge is then used in the product development process. We demonstrate, how production and marketing agents coordinate their actions in order to produce optimal products with respect to their incentive schemes. Our simulation shows that coordinating incentive schemes increase the performance of the firm. For a given incentive scheme, the House of Quality approach always outperformed the trial and error search. An interesting feature of the HoQ approach lies in the fact that product improvement is considerably faster compared to the alternative search strategy. (author's abstract) / Series: Working Papers SFB "Adaptive Information Systems and Modelling in Economics and Management Science"

Zugl.: Kiel, Univ., Diss., 2005