Comparison between different modeling techniques for assessing the role of environmental variables in predicting the catches of major pelagic fishes off India’s north-west coast
{"title":"Comparison between different modeling techniques for assessing the role of environmental variables in predicting the catches of major pelagic fishes off India’s north-west coast","authors":"V. K. Yadav, S. Jahageerdar, J. Adinarayana","doi":"10.56042/ijms.v51i02.66010","DOIUrl":null,"url":null,"abstract":"The contribution of four variables, namely Chlorophyll -a (Chl- a ), Sea Surface Temperature (SST), diffuse attenuation coefficient (Kd_490 or Kd), and Photosynthetically Active Radiation (PAR), in predicting the catches of major pelagic fish species (Indian mackerel, horse mackerel, Bombay duck, oil sardine, and other sardines) was evaluated using Canonical Correlation Analysis (CCA). The outcome of the analysis was compared with those obtained by using the following models and methods: the Generalized Linear Model (GLM), the Generalized Additive Model (GAM), connection weight methods, and the explanatory methods of Artificial Neural Networks (ANNs). Both the sets of results were in agreement. Neither the GAM nor the ANNs method showed any clear advantage over each other, although the GAM performed better than the GLM.","PeriodicalId":51062,"journal":{"name":"Indian Journal of Geo-Marine Sciences","volume":"47 1","pages":""},"PeriodicalIF":0.5000,"publicationDate":"2022-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Indian Journal of Geo-Marine Sciences","FirstCategoryId":"89","ListUrlMain":"https://doi.org/10.56042/ijms.v51i02.66010","RegionNum":4,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q4","JCRName":"Earth and Planetary Sciences","Score":null,"Total":0}
引用次数: 0
Abstract
The contribution of four variables, namely Chlorophyll -a (Chl- a ), Sea Surface Temperature (SST), diffuse attenuation coefficient (Kd_490 or Kd), and Photosynthetically Active Radiation (PAR), in predicting the catches of major pelagic fish species (Indian mackerel, horse mackerel, Bombay duck, oil sardine, and other sardines) was evaluated using Canonical Correlation Analysis (CCA). The outcome of the analysis was compared with those obtained by using the following models and methods: the Generalized Linear Model (GLM), the Generalized Additive Model (GAM), connection weight methods, and the explanatory methods of Artificial Neural Networks (ANNs). Both the sets of results were in agreement. Neither the GAM nor the ANNs method showed any clear advantage over each other, although the GAM performed better than the GLM.
期刊介绍:
Started in 1972, this multi-disciplinary journal publishes full papers and short communications. The Indian Journal of Geo-Marine Sciences, issued monthly, is devoted to the publication of communications relating to various facets of research in (i) Marine sciences including marine engineering and marine pollution; (ii) Climate change & (iii) Geosciences i.e. geology, geography and geophysics. IJMS is a multidisciplinary journal in marine sciences and geosciences. Therefore, research and review papers and book reviews of general significance to marine sciences and geosciences which are written clearly and well organized will be given preference.