The Visolvia technology
"Fully Automatic mode"
Visolvia's research within the areas of machine learning, signal processing and Neuroeconomics has resulted in totally unique technologies for Artificial Intelligence (AI) that have been incorporated into the financial IT-system Visolvia TradeRunner. The system is using AI both to produce new algorithms and to select which ones to use for the portfolio. Visolvia TradeRunner therefore replaces both the R&D department and the human decision makers required to operate the traditional algo-trading products.
Removing the humans in the loop makes it possible to carry out long term back-testing of asset management portfolios. And to ensure that uncertainties are kept within known limits Visolvia focuses its research on very low frequency trading systems.
Visolvia's technology is constructed from some basic elements that are able to filter information and make decisions at the same time; there is no need to separate these functions. The elements are of a few different types with somewhat different roles, but they all share a similar construction. The elements are interconnected and arranged in clusters, which in turn are connected in higher level clusters and so on. Thanks to the architecture Visolvia TradeRunner can process large amount of data in parallel.
The system automatically adjusts how much weight it should put on different sources of information and how elements and connections should be configured for a certain task. The way it deals with decision making is a bit fuzzy rather than precise, but the system is able to use its idle time to plan strategies that can be acted on very quickly when needed.
Neuro technologyIn some aspects the Visolvia technology reminds of biological neurons and how they are combined and used in nature, but the technology is not of the type of neural network technologies described in computer science literature. The principles are completely different and unlike traditional neural network computer technologies the Visolvia technology is extremely immune to overtraining and unknown states.
Presently Visolvia Trade-Runner is implemented using software on standard computer servers, but the basic elements could in principle be implemented in a number of different types of technologies, e.g. optical for extremely high speed computational power, or for other applications (such as complementing smart materials for e.g. intelligent clothing or biometrics) using nanotechnology and/or e.g. organic chemistry for simplified mass production.
Photo: Slowly but surely; half a billion years of proven neural technology.