Biochemical Oxygen Demand Prediction Based on Three-Dimensional Fluorescence Spectroscopy and Machine Learning DOI Creative Commons
Xu Zhang, Yihao Zhang,

Xuanyi Yang

et al.

Sensors, Journal Year: 2025, Volume and Issue: 25(3), P. 711 - 711

Published: Jan. 24, 2025

Biochemical oxygen demand (BOD) is an important indicator of the degree organic pollution in water bodies. Traditional methods for BOD5 determination, although widely used, are complicated and dependent on accurate chemical measurements dissolved oxygen. The aim this study was to propose a facile method predicting biochemical by fluorescence signals using three-dimensional spectroscopy parallel factor analysis combination with machine learning algorithm. samples were incubated five days national standard method, during which contents data measured at eight-hour intervals. maximum intensity three components decomposed extracted analysis. relationship between values established random forest model. results showed that there good correlation BOD values. effectively predicted model high goodness fit (R2 = 0.878) low mean square error (MSE 0.28). Although did not shorten incubation time, successful prediction realized non-contact measurement signals. This avoids operation DO improves detection efficiency, provides convenient solution analyzing large quantities monitoring quality.

Language: Английский

Artificial Intelligence and/or Machine Learning Algorithms in Microalgae Bioprocesses DOI Creative Commons
Esra İmamoğlu

Bioengineering, Journal Year: 2024, Volume and Issue: 11(11), P. 1143 - 1143

Published: Nov. 13, 2024

This review examines the increasing application of artificial intelligence (AI) and/or machine learning (ML) in microalgae processes, focusing on their ability to improve production efficiency, yield, and process control. AI/ML technologies are used various aspects such as real-time monitoring, species identification, optimization growth conditions, harvesting, purification bioproducts. Commonly employed ML algorithms, including support vector (SVM), genetic algorithm (GA), decision tree (DT), random forest (RF), neural network (ANN), deep (DL), each have unique strengths but also present challenges, computational demands, overfitting, transparency. Despite these hurdles, shown significant improvements system performance, scalability, resource well cutting costs, minimizing downtime, reducing environmental impact. However, broader implementations face obstacles, data availability, model complexity, scalability issues, cybersecurity threats, regulatory challenges. To address solutions, use simulation-based data, modular designs, adaptive models, been proposed. contributes literature by offering a thorough analysis practical applications, benefits critical insights into this fast-evolving field.

Language: Английский

Citations

8

Research on water quality detection integrating spectral analysis and automated control DOI

Xiaoman Huang,

Juntao Xiong,

H. W. Lin

et al.

Spectrochimica Acta Part A Molecular and Biomolecular Spectroscopy, Journal Year: 2025, Volume and Issue: 339, P. 126260 - 126260

Published: April 17, 2025

Language: Английский

Citations

0

Emerging biomedical applications of surface-enhanced Raman spectroscopy integrated with artificial intelligence and microfluidic technologies DOI

Zehra Taş,

Fatih Çiftçi, Kutay İçöz

et al.

Spectrochimica Acta Part A Molecular and Biomolecular Spectroscopy, Journal Year: 2025, Volume and Issue: 339, P. 126285 - 126285

Published: April 23, 2025

Language: Английский

Citations

0

Machine Learning Advancements and Strategies in Microplastic and Nanoplastic Detection DOI
Lifang Xie,

Minglu Ma,

Qiuyue Ge

et al.

Environmental Science & Technology, Journal Year: 2025, Volume and Issue: unknown

Published: April 28, 2025

Microplastics (MPs) and nanoplastics (NPs) present formidable global environmental challenges with serious risks to human health ecosystem sustainability. Despite their significance, the accurate assessment of MP NP pollution remains hindered by limitations in existing detection technologies, such as low resolution, substantial data volumes, prolonged imaging times. Machine learning (ML) provides a promising pathway overcome these enabling efficient processing complex pattern recognition. This systematic Review aims address gaps examining role ML techniques combined spectroscopy improving characterization NPs. We focused on application key tools detection, categorizing literature into aspects: (1) Developing tailored strategies for constructing models optimize plastic while expanding monitoring capabilities. Emphasis is placed harnessing unique molecular fingerprinting capabilities offered spectroscopy, including both infrared (IR) Raman spectra. (2) Providing an in-depth analysis issues encountered current approaches detection. highlights critical advancing our further, deeper investigation widespread presence By identifying challenges, this valuable insights future direction management public protection.

Language: Английский

Citations

0

Biochemical Oxygen Demand Prediction Based on Three-Dimensional Fluorescence Spectroscopy and Machine Learning DOI Creative Commons
Xu Zhang, Yihao Zhang,

Xuanyi Yang

et al.

Sensors, Journal Year: 2025, Volume and Issue: 25(3), P. 711 - 711

Published: Jan. 24, 2025

Biochemical oxygen demand (BOD) is an important indicator of the degree organic pollution in water bodies. Traditional methods for BOD5 determination, although widely used, are complicated and dependent on accurate chemical measurements dissolved oxygen. The aim this study was to propose a facile method predicting biochemical by fluorescence signals using three-dimensional spectroscopy parallel factor analysis combination with machine learning algorithm. samples were incubated five days national standard method, during which contents data measured at eight-hour intervals. maximum intensity three components decomposed extracted analysis. relationship between values established random forest model. results showed that there good correlation BOD values. effectively predicted model high goodness fit (R2 = 0.878) low mean square error (MSE 0.28). Although did not shorten incubation time, successful prediction realized non-contact measurement signals. This avoids operation DO improves detection efficiency, provides convenient solution analyzing large quantities monitoring quality.

Language: Английский

Citations

0