Clusters Patterning in High Tech Stock Market
Authors: | Tomarchio Giuseppina, Universita Degli Studi di Catania, Italy Bucolo Maide, Universita Degli Studi di Catania, Italy Galvagno Luca, Universita Degli Studi di Catania, Italy Fortuna Luigi, Universita Degli Studi di Catania, Italy |
---|
Topic: | 9.1 Economic & Business Systems |
---|
Session: | Finance and Banking |
---|
Keywords: | Financial Systems, Time series Classification, Neural Networks |
---|
Abstract
Stock Markets, like many other natural systems, are complex systems in a state of unstable equilibrium where future evolution depends on many parameters. Their evolution is characterized by different factors: political and social events, government policies, humour and strategy of the consumer environment. Even if attention has been focused on this system since the early years of the nineteen-century and many efforts are still concentrated on it, no valid and recognized model has been accepted by the scientific community. This paper aims at being a scientific contribution which tries to formalize the main features in the stock market topology. Cluster formation has been investigated in different time windows to identify specific properties. The analysis has been concluded in this case to highlight clusters in industrial sectors linked to the semiconductor industry.