An Enhanced Extreme Learning Machine for Dissolved Oxygen Prediction in Wireless Sensor Networks

Abstract

Water quality monitoring using Wireless Sensor Networks (WSNs) is essential in aquaculture water quality management. In the field of water quality monitoring, dissolved oxygen (DO) is a key parameter, and its prediction can provide decision support for aquaculture production, thereby reducing farming risk. However, it is difficult to build a precise prediction model, and existing methods of DO prediction neglect the importance of analyzing DO content. To address this problem, this study proposes a hybrid DO prediction model, named KIG-ELM, which is composed of K-means, improved genetic algorithm (IGA), and extreme learning machine (ELM). This model is based on edge computing architecture, in which data acquisition, processing and dissolved oxygen prediction are distributed in sensing nodes, routing nodes and server respectively. Sensing technique and clustering operation are applied in the process of data acquisition and processing. Meanwhile, an optimized extreme learning machine is implemented for DO prediction. We evaluate the efficiency and accuracy of proposed prediction approach in a practical aquaculture on massive water quality data. Experimental results show that the hybrid model achieves significant prediction results and can meet the needs of practical production and management.

A CNN-OSELM Multi-Layer Fusion Network With Attention Mechanism for Fish Disease Recognition in Aquaculture

“…It has been experimented that random addition of hidden layer nodes did not guarantee better performance of ELM [54]. The recommendation is to automatically determine the initial hidden layer node number by using incremental constructive techniques such as incremental extreme learning machine (I-ELM) [51] and enhanced incremental ELM (EI-ELM) [14], among others and then optimizing the initial model in order to obtain optimal network. In this study we adopt the number of hidden nodes as 4096 because it has been experimented on similar output size in [54] and proven to perform well.…”

“…Existing studies on indirect approaches to fish disease identification include detection of anomalous behavior [11], prediction of water quality [12] and dissolved oxygen [13], [14], [15], [16], [17]. The computer vision-based direct approaches to fish disease detection utilize the physical parameters or physical appearance of fish such as body texture, eye color, appearance of fish head, fish fins, scales, gills and tail to recognize disease.…”

Simultaneous Detection of Multiple Surface Acoustic Wave Sensor Tags for Water Quality Monitoring Utilizing Cellular Code-Reuse Approach

“…Several work has been carried out in the area of WSNs for water quality monitoring among which paper [19] presents a smart low-cost remote monitoring sensor and utilizes it for performing real-time water quality monitoring. Paper [20] describes a multi-parameter DO prediction model utilizing a collection of water quality sensors and paper [21] presented an IoT platform which was applied to conduct water contaminant detection case studies. With reference to the work in SAW tags, paper [22] proposed various anti-collision strategies for SAW tags to improve the detectability.…”

“…Accordingly, the sensor-tags were not needed to be bought and physically tested. In the light of the above, the scope and contributions of our paper are as follows: Comparing with papers [19][20][21][22][23][24][25][26][27], we present a novel and efficient approach to water quality monitoring using cellular code-reuse and to simultaneously identify, and, implicitly, sense multiple orthogonal SAW sensor-tags. The so-developed approach is intended to monitor the water quality over a wide-area pond, lakes, rivers, and coastal sea where many water-based assets and other resources may lie.…”

IOTA-Based Mobile Crowd Sensing: Detection of Fake Sensing Using Logit-Boosted Machine Learning Algorithms

“…In research, both ML and DL techniques have been proven to extract meaningful patterns from network data to classify flows as anomalous or benign. Logit-boosted algorithms has demonstrated speed in learning valuable characteristics from raw data and has emphasized integrating IOTA networks into NIDS [9]. Logit-boosted algorithm is a machine learning approach researched by academics in data mining, data science, and network security [10].…”