Prediction Model of Ammonia Nitrogen Concentration in Aquaculture Based on Improved AdaBoost and LSTM

Mathematics, Год журнала: 2024, Номер 12(5), С. 627 - 627

Опубликована: Фев. 20, 2024

The concentration of ammonia nitrogen is significant for intensive aquaculture, and if the too high, it will seriously affect survival state aquaculture. Therefore, prediction control in advance essential. This paper proposed a combined model based on X Adaptive Boosting (XAdaBoost) Long Short-Term Memory neural network (LSTM) to predict mariculture. Firstly, weight assignment strategy was improved, number correction iterations introduced retard shortcomings data error accumulation caused by AdaBoost basic algorithm. Then, XAdaBoost algorithm generated several LSTM su-models concentration. Finally, there were two experiments conducted verify effectiveness model. In experiment, compared with other comparison models, RMSE XAdaBoost–LSTM reduced about 0.89–2.82%, MAE 0.72–2.47%, MAPE 8.69–18.39%. stability RMSE, MAE, decreased 1–1.5%, 0.7–1.7%, 7–14%. From these experiments, evaluation indexes superior which proves that has good accuracy lays foundation monitoring regulating change future.

Язык: Английский

---

Seasonal variation of the quality of groundwater resources for human consumption and industrial purposes in the central plain zone of Punjab, India

Environmental Monitoring and Assessment, Год журнала: 2023, Номер 195(12)

Опубликована: Ноя. 11, 2023

Язык: Английский

---

Enhancing river health monitoring: Developing a reliable predictive model and mitigation plan

Ecological Indicators, Год журнала: 2023, Номер 156, С. 111190 - 111190

Опубликована: Ноя. 6, 2023

The escalating environmental harm inflicted upon rivers is an unavoidable outcome resulting from climate fluctuations and anthropogenic activities, leading to a catastrophic impact on water quality thousands of individuals succumb waterborne diseases. Consequently, the monitoring stations have been established worldwide. Regrettably, real-time evaluation Water Quality Index (WQI) hindered by intricate nature off-site parameters. Thus, there pressing need create precise robust prediction model. dynamic non-linear characteristics parameters pose significant challenges for conventional machine learning algorithms like multi-linear regression, as they struggle capture these complexities. In this particular investigation, model called Feedforward Artificial Neural Networks (FANNs) was employed develop WQI Batu Pahat River, Malaysia exclusively utilizing on-site proposed method involves consideration whether include or exclude such BOD COD, which are not measured in real time can be costly monitor inputs. Validation accuracy values 99.53%, 97.99%, 91.03% were achieved three different scenarios: first scenario utilized full input, second excluded BOD, third both COD. It suggested that has better predictive power between input variables output variables. Factor contributed river pollution identified mitigation plan proposed. This could provide effective alternative compute pollution, manage resources mitigate negative impacts change ecosystems.

Язык: Английский

---

Assessing Groundwater Quality for Sustainable Drinking and Irrigation: A GIS-Based Hydro-Chemical and Health Risk Study in Kovilpatti Taluk, Tamil Nadu

Water, Год журнала: 2023, Номер 15(22), С. 3916 - 3916

Опубликована: Ноя. 9, 2023

The continuous investigation of water resources is essential to assess pollution risks. This study investigated a groundwater assessment in the coastal belt Tamil Nadu’s Kovilpatti Taluk, Thoothukudi district. Twenty-one samples were collected during pre-monsoon and post-monsoon seasons, analyzing quality parameters, namely pH, EC, Cl−, SO42−, Ca2+, Mg2+, HCO3−, TH, Na2+, K+. Water Quality Index (WQI) was computed it observed that 5% 9% unsuitable for drinking. SAR, MHR, RSC, %Na Kelley’s index used determine irrigation suitability. Pre-monsoon shows 29% (MHR) 71% (RSC) unsuitable, 59% unsuitable. Coastal activity, urbanization, industrialization resulted degradation quality. Solving this issue requires sustainable wastewater treatment strict industrial discharge guidelines. Spatial distribution plots, Box Gibbs Piper Wilcox plots Correlation Matrices had similar results WQI its physical–chemical parameters. According human health risk assessment, Mooppanpatti, Illuppaiurani, Vijayapuri regions show high risks due nitrate fluoride concentration groundwater. Kadambu, Melparaipatti, Therkuilandhaikulam, Vadakku Vandanam have low levels, posing minimal risk.

Язык: Английский

---

Predicting water quality through daily concentration of dissolved oxygen using improved artificial intelligence

Scientific Reports, Год журнала: 2023, Номер 13(1)

Опубликована: Ноя. 21, 2023

As an important hydrological parameter, dissolved oxygen (DO) concentration is a well-accepted indicator of water quality. This study deals with introducing and evaluating four novel integrative methods for the prediction DO. To this end, teaching-learning-based optimization (TLBO), sine cosine algorithm, cycle algorithm (WCA), electromagnetic field (EFO) are appointed to train commonly-used predictive system, namely multi-layer perceptron neural network (MLPNN). The records USGS station called Klamath River (Klamath County, Oregon) used. First, networks fed by data between October 01, 2014, September 30, 2018. Later, their competency assessed using belonging subsequent year (i.e., from 2018 2019). reliability all models, as well superiority WCA-MLPNN, was revealed mean absolute errors (MAEs 0.9800, 1.1113, 0.9624, 0.9783) in training phase. calculated Pearson correlation coefficients (RPs 0.8785, 0.8587, 0.8762, 0.8815) plus root square (RMSEs 1.2980, 1.4493, 1.3096, 1.2903) showed that EFO-MLPNN TLBO-MLPNN perform slightly better than WCA-MLPNN testing Besides, analyzing complexity time pointed out most efficient tool predicting In comparison relevant previous literature indicated suggested models provide accuracy improvement machine learning-based DO modeling.

Язык: Английский

---