Thesis (MBA)--Stellenbosch University, 1999. / ENGLISH ABSTRACT: Neural networks are computer systems that attempt to mimic the operation of the human brain. In contrast to traditional systems these systems can learn and will change their behaviour over time. In the highly competitive business environment of today, neural networks is one of many technologies that can assist organisations in gaining a competitive advantage.
Neural networks also find application in the financial services industry. Applications range from corporate distress or failure models to forecasting of stock prices and many others. Generally speaking, neural networks often offer an exciting alternative to traditional methods of forecasting and classification in this industry. Neural networks must be implemented with care and judgement, as their performance is sensitive with respect to their construction and architecture.
Neural networks, as with other technologies, rarely operate in isolation. Neural networks can be integrated with expert systems, genetic algorithms, data mining and even traditional statistical and operational research techniques. Integration produces systems in which the whole is greater than the sum of the parts.
Neural networks are also researched and applied in the South African financial services industry, both at an academic and commercial level. Indications are that South Africa is not far behind the international community in exploring the benefits of neural networks. / AFRIKAANSE OPSOMMING: Neurale netwerke is rekenaarstelsels wat poog om die werking van die menslike brein na te boots. In kontras met tradisionele stelsels, leer neurale netwerke en verander dus hul gedrag met verloop van tyd. In vandag se hoogs kompeterende besigheids omgewing, is neural netwerke een van vele tegnologieë wat organisasies kan gebruik om ‘n mededingende voordeel te bekom.
Neurale netwerke het ook toepassing in die finansiële dienste industrie. Toepassings wissel van korporatiewe mislukkings modelle tot die vooruitskatting van aandele pryse en vele ander. Neurale netwerke bied ‘n opwindende alternatief tot tradisionele modelle vir vooruitskatting en klassifikasie. Toepassings van neurale netwerke moet egter met oorleg plaasvind, aangesien hul prestasie sterk afhanklik is van hul konstruksie en argitektuur.
Soos met ander tegnologie, word neurale netwerke selde in isolasie geïmplementeer. Neurale netwerke kan met sukses geïntegreer word met ekspert stelsels, genetiese algoritmes, data ontginnings metodes sowel as tradisionele statistiese of operasionele navorsings metodes. Integrasie bied stelsels wat meer bied as die som van die onafhanklike komponente.
Neurale netwerke word ook in die Suid-Afrikaanse finansiële industrie nagevors en toegepas. Alle indikasies dui daarop dat, met betrekking tot die navorsing van voordele van neurale netwerke, Suid Afrika nie ver agter die internasionale gemeenskap is nie.
| Identifer | oai:union.ndltd.org:netd.ac.za/oai:union.ndltd.org:sun/oai:scholar.sun.ac.za:10019.1/85178 |
| Date | 12 1900 |
| Creators | Van Wamelen, Riaan Joop |
| Contributors | Gevers, W. R., Stellenbosch University. Faculty of Economic and Management Sciences. Graduate School of Business. |
| Publisher | Stellenbosch : Stellenbosch University |
| Source Sets | South African National ETD Portal |
| Language | en_ZA |
| Detected Language | English |
| Type | Thesis |
| Format | viii, 77 p. : ill. |
| Rights | Stellenbosch University |
Page generated in 0.0019 seconds