2023 Early Hearing Detection & Intervention Conference
March 5-7, 2023 • Cincinnati, OH
6/05/2017 | 3:00 PM - 3:15 PM | EVOLVING BETTER INDICES: CAN GENETIC ALGORITHMS IMPROVE MULTIMETRIC INDEX CONSTRUCTION? | 302C
EVOLVING BETTER INDICES: CAN GENETIC ALGORITHMS IMPROVE MULTIMETRIC INDEX CONSTRUCTION?
For the last four decades, multimetric indices have remained a standard tool in quantitative aquatic bioassessment. During this time a set of construction guidelines has emerged to ensure that indices reliably portray the degree to which a waterbody’s biological condition differs from regional reference streams. Many steps in this standardized process rely on subjective judgment and do not ensure that the final index is better than all possible candidate indices. Here I explore how optimization methods mimicking natural selection might eliminate these shortcomings. The particular metrics included, the level to which taxa are identified and scoring routines for metrics act as genes in the DNA of any multimetric index, the fitness of which is the index’s ability to discriminate between reference and disturbed streams. In theory then, a genetic algorithm can evolve indices with higher discrimination ability. After comparing the performance of a regionalized index constructed with a genetic algorithm to one constructed using standard methods, I report the synergistic improvements evolutionary programming and ecological understanding can bring to the index construction process. Finally, I highlight future research directions for improving aquatic bioassessment methods using the evolutionary programming toolbox.
- C12 Conservation Ecology
- C36 Water Resource Management
- S31 Moving forward in flow ecology: identifying and testing key hypotheses & C17 Bioassessment
Presentation:
This presentation has not yet been uploaded.
Handouts:
Handout is not Available
Transcripts:
CART transcripts are NOT YET available, but will be posted shortly after the conference
Presenters/Authors
Kristofor Voss
(), Regis University, kvoss@regis.edu;
ASHA DISCLOSURE:
Financial -
Nonfinancial -