Application of Extreme Learning Machine for Predicting Chlorophyll‐a Concentration Inartificial Upwelling Processes

Abstract

Artificial upwelling, artificially pumping up nutrient-rich ocean waters from deep to surface, is increasingly applied to stimulating phytoplankton activity. As a proxy for the amount of phytoplankton present in the ocean, the concentration of chlorophyll a (chl-a) may be influenced by water physical factors altered in artificial upwelling processes. However, the accuracy and convenience of measuring chl-a are limited by present technologies and equipment. Our research intends to study the correlations between chl-a concentration and five water physical factors, i.e., salinity, temperature, depth, dissolved oxygen (DO), and pH, possibly affected by artificial upwelling. In this paper, seven models are presented to predict chl-a concentration, respectively. Two of them are based on traditional regression algorithms, i.e., multiple linear regression (MLR) and multivariate quadratic regression (MQR), while five are based on intelligent algorithms, i.e., backpropagation-neural network (BP-NN), extreme learning machine (ELM), genetic algorithm-ELM (GA-ELM), particle swarm optimization-ELM (PSO-ELM), and ant colony optimization-ELM (ACO-ELM). These models provide a quick prediction to study the concentration of chl-a. With the experimental data collected from Xinanjiang Experiment Station in China, the results show that chl-a concentration has a strong correlation with salinity, temperature, DO, and pH in the process of artificial upwelling and PSO-ELM has the best overall prediction ability.

Methods and Models for Electric Load Forecasting: A Comprehensive Review

“…Further discussion and details about the ELM networks and their more advanced variants can be found in [71][72][73][74][76][77][78][79][80]. There, a more detailed explanation about efficient using of the ELM networks for the purpose of ELF prediction can be found as well.…”

A Self-adaptive differential evolutionary extreme learning machine (SaDE-ELM): a novel approach to blast-induced ground vibration prediction

“…where each element x i is a single column matrix of length p which is equivalent to the total number of datasets as shown in Eq. (37).…”

“…In that regard, Huang et al [33] proposed the kernel-based ELM (KELM) which adds a positive regularisation coefficient in the computation of the output weight to enhance stability and a kernel matrix when the hidden layer feature mapping is unknown. Additionally, other researchers tend to rely on the search abilities of metaheuristic optimisation algorithms notably particle swarm optimisation (PSO) and genetic algorithm (GA) to select the optimal input weights and hidden neurons of the ELM approach [34][35][36][37]. With regard to blasting studies, authors such as Armaghani et al [38] developed a novel hybrid ELM model optimised by autonomous groups particle swarm optimisation (AGPSO) algorithm for blastinduced ground vibration prediction.…”

Comparing the performance of machine learning algorithms for remote and in situ estimations of chlorophyll‐a content: A case study in the Tri An Reservoir, Vietnam

“…The field‐based approach is benefitting of active sampling and synchronizing with in situ Chl‐a measurement, and potentially in higher accuracy of retrieval model. However, this approach might be spatial limited and less rationale in the case of large‐scale mapping (Toming et al, 2016; Wei et al, 2019). The remote sensing approach, in contrast, provides a more conventional approach to deal with the large‐scale mapping of severe blooms and multitemporal monitoring.…”