作者: Michael Meyners
DOI: 10.1016/J.FOODQUAL.2018.11.022
关键词: Artificial intelligence 、 Machine learning 、 Selection (genetic algorithm) 、 Arbitrariness 、 Discriminative model 、 Modality (human–computer interaction) 、 Dominance (ethology) 、 Computer science
摘要: Abstract A workshop held at the 2018 Sensometrics conference in Montevideo aimed a data-based comparison between three temporal methods, namely Temporal Check-All-That-Apply (TCATA), Dominance of Sensations (TDS), and by Modality (MTDS). Data on 8 yoghurts designed following full-factorial design have been evaluated using protocols. In this paper, we compare data obtained with regard to discriminative power highlight potential pitfalls when making significance-based comparisons. TCATA is found provide overall highest discrimination, while for special purposes, other methods may advantages. We further dominance selection rates from addressing arbitrariness necessary standardizations. Finally, results correspondence analyses trajectories, showing noteworthy differences discussion, it argued that reasonable might not be possible or even reasonable, as they address quite different concepts. Rather than being considered alternatives same objective, should better complimentary, project needs.