Sajjad Haider, Samad Hassan, and Mohammad Nishat
Influence Nets, Bayesian Networks, Knowledge Engineering, Financial Informatics
This paper demonstrates how Influence Nets can aid in strategic level decision making about investment in a stock market. Unlike most of the reported techniques, which look at the past micro-level data to develop a predictive model, the analysis done via Influence Nets work at the macro or strategic level and helps in identifying the main factors that impact the variable(s) of interest. In addition, it also caters well for unprecedented or not-so-frequent events whose impacts on the decision variable(s) are not captured appropriately in the past data but is quite clear in the mind of subject matter experts. Influence Nets allow subject matter experts to define cause and effect relationships among variables within a problem domain. In addition, with the aid of tools like sensitivity and what-if analyses it helps in identifying the most important factors affecting variable(s) of interest.
Important Links:
Go Back
