Scope
Our goal with the Feature Generation module is to provide a knowledge-intensive and
computationally efficient coarse-grained analysis of historical prices which can be analyzed
further in a second layer of reasoning. The domain knowledge implemented in the module
is thus limited to methods and techniques in technical analysis. The technical analysis
literature includes a wealth of different stock analysis techniques, some of which involve
complicated and intricate price patterns subjective in both detection and interpretation.
These methods would be both computationally expensive to detect and evaluate, and have
consequently been disregarded. We thus apply Occam’s razor to the choice of methods
in technical analysis, focusing on the most popular indicators that can be efficiently
operationalized and are intuitive in interpretation.
It may seem overly presumptuous to believe that historical price fluctuations alone can
be used to predict the direction of future prices. It may thus seem natural to include
some fundamental analysis knowledge in the feature generation process. However, due to
the inherent limitations in time and the added complexity of including a second analysis
technique, this has not been a priority. We have instead placed focus on creating a model
that can be easily extended with new analysis techniques, not necessarily from technical
analysis, by using two separate reasoning layers and using an agent-oriented approach for
the domain knowledge. The agent-oriented approach is explained in detail in Chapter 3.
Although our goal in this thesis is not to justify, prove or disprove technical analysis, by
focusing strictly on technical indicators we are presented with an opportunity to evaluate
the utility of selected methods in this form of stock analysis.