Genetic Programming: Theory and Practice

Diagnosing Faults in Power Transformers With Variational Autoencoder, Genetic Programming, and Neural Network

“…GP is usually applied to a mapping problem of a set of inputs in an output dataset, and for this purpose, the algorithm attempts to perform the synthesis of functions formed by linear and nonlinear mathematical operators that best relate the input/output correspondence [63].…”

Genetic programming theory and practice: a fifteen-year trajectory

“…2 From I to XV GPTP I: 2003. Worzel and Riolo [19] wrote in the inaugural volume: "The dynamics of the GP process is poorly understood with many serious questions remaining. People applying GP to real-world problems have relied more on intuition than theory, experience more than mathematics.…”

“…In the first workshop there seemed to be a rather wide divide between theoreticians and practitioners, the former aiming for simple models that could be analyzed thoroughly, the latter content with viewing success as its own justification [19].…”

“…GPTP X: 2012. Riolo et al [9] pointed out an important change in the tenth edition of the workshop: a more varied mix of different representations of GP individuals, e.g., SQL query, image filter, power control algorithm, and gameboard evaluation function. Moreover, a conscious effort was made to go beyond GP symbolic regression, an application that starred in previous events.…”

“…Riolo et al [9] remarked on the increasing gap between theory and practice, expressing doubt it would ever be closed. In addition to necessary continued research into theory, several other directions were raised by the participants, including: what problems can be solved using a computer program and how can GP be used therein; how to seamlessly integrate different types of data structures; various requirements for open-ended evolution; exploiting modern architectures to run GP; post-processing and selection of GP solutions.…”