Method and system for the analysis of the cause of marine algae and its concentration prediction based on machine learning

Abstract

The invention discloses a method and system for analyzing the cause of marine algae and predicting its concentration based on machine learning. The method includes the following steps: constructing a data set, performing standardized processing on it, and dividing the processed data set into a training set and a test set set; feature selection is performed on the environmental parameters in the training set to obtain a variety of feature subsets, and all feature subsets are verified multiple times on a variety of different machine learning algorithms to obtain the optimal feature subset corresponding to each machine learning algorithm and its evaluation index; select the machine learning algorithm with the best evaluation index as the optimal prediction model; use the optimal prediction model to predict the algae concentration corresponding to the optimal feature subset; use the GBDT model to train the environmental parameter data in the data set, and obtain the optimal The importance of each environmental parameter in the optimal feature subset is used to analyze the cause of algal formation.

Description

technical field [0001] The present disclosure relates to the technical field of harmful algal bloom prediction, in particular to a machine learning-based method and system for analyzing the cause of marine algae and predicting its concentration. Background technique [0002] In the research on the prediction of harmful algal blooms, the existing prediction methods of harmful algal blooms mainly include: nonlinear dynamics, statistical prediction methods, machine learning prediction methods, etc. [0003] The nonlinear dynamics method theoretically studies the ecological dynamic behavior of single-population red tide algae, multi-population red tide algae and red tide food chain, clarifies the nonlinear dynamic characteristics of the model, and puts forward the ecological dynamic mechanism of red tide occurrence. But it will be very difficult to solve when the number of features is large. [0004] Statistical forecasting methods are suitable for statistical analysis of large...

AI Technical Summary

Problems solved by technology

However, this method has certain requirements for the form of the sequence, which needs to meet the stationarity, and its prediction results have nothing to do with the environmental physical parameters, and cannot reflect the correlation

Images