Modeling method of fish habitat suitability index based on support vector machine

Abstract

The invention relates to a fish habitat suitability index modeling method based on a support vector machine, comprising the following steps: 1) combining multiple marine environmental factors to obtain a suitability index value SI corresponding to a group of marine factors; 2) Normalize all marine environmental factors and SI; 3) Based on the data in step 2), establish a support vector machine HSI model: SVM HSI; 4) Use SVM HSI to predict different areas and obtain the root mean square error of the prediction and correlation coefficients, and forecast the fishery situation in the GIS environment; 5) output and save the simulation results. Compared with the prior art, when the number of samples is small, the present invention can better excavate the laws existing in the data, and obtain the advantages of a more reasonable HSI model and fish situation forecast results.

Description

technical field [0001] The invention relates to a fish habitat index modeling method, in particular to a fish habitat suitability index modeling method based on a support vector machine. Background technique [0002] The Habitat Suitability Index (HSI) model was first proposed by American scientists in the 1980s to simulate the response of organisms to the surrounding habitat elements. At present, it has been widely used in the fields of species distribution and management, and has gradually been applied in the analysis and prediction of fish fisheries, and achieved good results. The research of HSI is not only to theoretically understand the relationship between fishery distribution and marine environmental elements, but also to provide information reference for fishing production and fishery resource managers. The HSI model has been applied to fishery analysis of various fish species, such as bigeye tuna in the Indian Ocean, squid in the Atlantic and Pacific Oceans, macke...

AI Technical Summary

Problems solved by technology

[0003] There is a dynamic interactive relationship between marine environmental elements and the central fishery and constitutes a complex system. There is usually a certain correlation between environmental elements. However, traditional mathematical statistical methods cannot eliminate the inherent multiple correlations of environmental elements when constructing an HSI model. , the accuracy of analysis and prediction is greatly limited

At the same time, due to the limitation of commercial fishing data, the number of sample points of fishery resource data is very limited to a large extent, and it is difficult to obtain effective models from it by ordinary methods

Images