Suomi
Gamma Exchange Model (GEM) is PsiFactor’s stock investing software that seeks for decision making models that seeks for good investment strategies to select good stock investment portfolios.
GEM creates stock investment strategies in a for of programming code and tests how well does the certain model fits into the data.
The model can be described simply in the next form:
1. The algorithm creates decision making models to select stock investment portfolios
2. Models are simulated and the algorithm checks how does the portfolios selected by the strategies behave with respect to known fitness function that measures return and risk using a certain metrics
3. The algorithm uses genetic optimization to optimize the decision making model. The basic operations are generating new strategy, mutation and crossover
GEM can be used to generate decision making models for variety type of instruments that has decent data available.
To avoid sample fit, GEM uses huge amount of processed stock data to find out what are good rules to select stock instruments. As a result of that, the algorithm consumes lots of resources and that is why it build using C/C++ and uses multi-threading.
GEM creates stock investment strategies in a for of programming code and tests how well does the certain model fits into the data.
The model can be described simply in the next form:
1. The algorithm creates decision making models to select stock investment portfolios
2. Models are simulated and the algorithm checks how does the portfolios selected by the strategies behave with respect to known fitness function that measures return and risk using a certain metrics
3. The algorithm uses genetic optimization to optimize the decision making model. The basic operations are generating new strategy, mutation and crossover
GEM can be used to generate decision making models for variety type of instruments that has decent data available.
To avoid sample fit, GEM uses huge amount of processed stock data to find out what are good rules to select stock instruments. As a result of that, the algorithm consumes lots of resources and that is why it build using C/C++ and uses multi-threading.