Enhanced Gene Expression Programming for signal-background discrimination in particle physics
Liliana Teodorescu, Zhengwen Huang · 2009
The original version of Gene Expression Programming, a variant of Evolutionary Algorithms, was enhanced in this study with an alternative representation of the candidate solution based on a prefix notation, and with a truncated evolution mechanism.The algorithm was applied to a signalbackground classification problem for which a dynamic classification threshold was implemented.As an example application the selection of K S particles produced in e + e -interactions at 10 GeV and reconstructed in the decay mode K S → π + π -was used.All these developments resulted in an algorithm more efficient in terms of its convergence speed, measured in number of generations, as well as in slight improvements of the quality of the final solution measured in terms of the classification accuracy.