Study On The Key Technology Of The Growth Model Of Semi-smooth Tongue Sole

Traditional semi-smooth tongue sole breeding relies mainly on experience and cannot meet the requirements of fine management of breeding production.Therefore,it is necessary to explore the relationship between growth factors of semi-smooth tongue sole,use machine learning and big data technology to build a growth model of semi-smooth tongue sole,monitor the growth and development of semi-smooth tongue sole,and realize refined management.This paper studies various key technologies in the modeling process,build a semi-smooth tongue sole growth model,and develops a semi-smooth tongue sole breeding decision support system based on this to achieve fine management and precise decisionmaking.The main study contents include:Aim at the problem of missing data in the semi-smooth tongue sole growth data collected under the factory-recirculating aquaculture model,two methods of adjacent point mean interpolation and linear interpolation were used to impute the missing data,and select the better method through comparison.In addition,in order to remove the influence of the data dimension on the model,the data is de-dimensionalized.Use correlation coefficient,KMO test,Bartlett sphere test and other methods to test the correlation of the data set.Use PCA,LDA and FA three methods to reduce the dimensionality of the data set,and build a BP neural network model for comparison,and draw a conclusion:the RMSE and MAPE of dimensionality reduction using the PCA method are smaller,and the dimensionality reduction effect is better.Aim at the problem of insufficient parallel ability of the current parallel BP neural network based on Map Reduce,a combined segmentation method is proposed to improve,and an improved parallel BP neural network based on Map Reduce is constructed to build a semismooth tongue sole growth model.Three methods of BP neural network,parallel BP neural network and improved parallel BP neural network are used to build the growth model of semismooth tongue sole for comparison,the results show that: the improved parallel BP neural network proposed in this paper has shorter running time,smaller RMSE and MAPE.On the basis of the growth model of semi-smooth tongue sole and the actual needs of enterprises,a decision support system for semi-smooth tongue sole breeding was developed to realize the fine management and precise decision-making of semi-smooth tongue sole breeding.