“…Given the analysis discussed above, we found six types of combinations which we can sort all listed publications ( Section 3 ) into (A table summing up these findings can be found in Appendix B ): - Complexity measures as an additional criterion for analysis: [ 7 , 16 , 17 , 19 , 20 , 21 , 22 , 23 , 24 , 25 , 26 , 27 , 28 , 29 , 30 , 31 , 32 , 33 ]
- Complexity measures to improve the architecture of the employed neural network/ algorithm: [ 21 , 23 ]
- Complexity measures as additional features for machine learning algorithms: [ 16 , 20 , 30 , 31 , 32 ]
- Complexity measures to find regions of increased predictability: [ 22 , 24 , 28 , 29 ]
- Feature Selection using Complexity Measures: [ 26 , 27 , 29 ]
- Filtering predictions/ensembles using Complexity Measures: [ 33 ]
…”