| The back testing procedures that we
use are meticulously scrutinized to ensure that the AI models have no unfair advantages.
When are always careful to avoid back testing on data that the
model has already been trained on. In order to properly back test
a model, one must only train the model on data that is available
up to the point in time that a prediction is made. For instance,
when back testing and making a prediction for July of 1992, the
model must be trained on historical data up to, but no later than,
June of 1992. Then, for the next month's predictions for August of
1992, the model must only be trained on data up till July of 1992.
This is very time consuming, but it is the right way to do it. Additionally, no optimization techniques were used to artificially boost results. In other words, we did not rerun the back testing over and over again, each time making small changes to the model until we achieved positive results. Optimizing in this way tends to make a model unstable. An unstable model can make the past results look spectacular, but if you try to use it for future predictions, it will perform poorly. |