Consolidation of Results
With reference to the experimental work in Chap. 4, we summarize the main findings in point form for the benefit of a wider readership. Notably, the fact that processing power and memory do not seem to have a significant effect on composing efficiency using the Digital Synaptic Neural Substrate (DSNS) approach, and how the quality of the compositions is higher compared to a random piece placement approach. In addition, the DSNS happens to be better than the present state-of-the-art ‘experience table’ approach that is also better at composing than simple random piece placement. Furthermore, in certain cases, variations in the number of photographs or chess games used to seed the DSNS process may be significant. Similarly, so do variations in the number of attributes used that represent the fragments of information taken from the aforemention domains. Possible limitations of the DSNS are discussed, including a brief exploration of its potential applications in other domains and fields.