Web interface generation and testing using artificial neural networks

摘要

Roughly described, the disclosed technique is based on artificial neural networks and evolutionary to efficiently identify the most successful web page designs in search space without testing all possible web page designs in search space. A so-called machine learning transformation optimization (MLCO) system that uses computation is provided. The search space is defined by the marketer based on the web page design. Neural networks are represented as genomes. Neural networks map user attributes from live user traffic to different dimensions and dimension values in the output funnel presented to the user in real time. The genome is subjected to evolutionary manipulations such as initialization, testing, competition, and reproduction to identify parent genomes that perform well and grandchildren that are likely to perform well. [Selection] Figure 2