Data Mining for Source Apportionment of Trace Elements in Water and Solid Matrix

Sampling and pre-treatment effects on the quantification of (nano)silver and selected trace elements in surface water - Application in a Dutch case study

This study reports pollution source apportionment of surface waters and human health risk assessment based on 18 physicochemical and traces elements from 24 water quality monitoring sites for surface and groundwater around the two trans-boundary rivers of Pakistan: The Ravi and Sutlej. The principal component analysis identified 6 principal components (76.98 % cumulative variance) which are mainly caused by untreated industrial effluents, intense agricultural activities, and irrigation tailwater discharges. For all dissolved trace elements in surface waters, health hazard indices (HI) and hazard quotients (HQ) through ingestion and dermal contact are < 1 except As and Cr through ingestion only (for both adults and children). For adults and children, the HQingestion and HI values for As, Mn, Cu (for children only) and As, Fe, Mn (for children only) are > 1, indicating that As, Mn, Fe, and Cu are the most important pollutants causing chronic risks among the selected trace elements in both shallow and deep groundwater respectively. HQingestion, HQdermal, and HI values are higher for children than that of adults which shows the high susceptibility of children to these dissolved trace elements. The carcinogenic indices for the entire surface water elements exceed 10−6 through dermal and ingestion pathways suggesting carcinogenic health risk to the surrounding community. Hence, to protect human health, wastewater treatment plants and best management practices should be practiced to control point source and nonpoint source pollution respectively in the understudied area.

Recent advances in computing, communications, digital storage technologies, and high-throughput data-acquisition technologies, make it possible to gather and store incredible volumes of data. It creates unprecedented opportunities for large-scale knowledge discovery from huge database. Data mining (DM) technology has emerged as a means of performing this discovery. There are countless researchers working on designing efficient data mining techniques, methods, and algorithms. Many data mining methods and algorithms have been developed and applied in a lot of application fields [1]. Unfortunately, most data mining researchers pay much attention to technique problems for developing data mining models and methods, while little to basic issues of data mining [2]. In this talk, some basic issues of data mining are addressed. What is data mining? What is the product of a data mining process? What are we doing in a data mining process? What is the rule we should obey in a data mining process? Through analyzing existing data mining methods, and domain-driven (or user-driven) data mining models [3-5], we find that we should take a data mining process as a process of knowledge transformation. Based on this understanding of data mining, a conceptual data mining model of domain-oriented data-driven data mining (3DM) is proposed [2]. The relationship between traditional domain-driven (or user-driven) data mining models and our proposed 3DM model is also analyzed. Some domain-oriented data-driven data mining algorithms for mining such knowledge as default rule [6], decision tree [7], and concept lattice [8] from database are proposed. The experiment results for these algorithms are also shown to illustrate the efficiency and performance of the knowledge acquired by our 3DM data mining algorithms.

Trace metal pollution poses a serious threat to the aquatic ecosystem. Therefore, characterizing the long-term environmental behavior of trace metals and their driving forces is essential for guiding water quality management. Based on a long-term data set from 1990 to 2019, this study systematically conducted the spatiotemporal trend assessment, influential factor analysis, and source apportionment of trace elements in the rivers of the German Elbe River basin. Results show that the mean concentrations of the given elements in the last 30 years were found in the order of Fe (1179.5 ± 1221 μg·L-1) ≫ Mn (209.6 ± 181.7 μg·L-1) ≫ Zn (52.5 ± 166.2 μg·L-1) ≫ Cu (5.3 ± 5.5 μg·L-1) > Ni (4.4 ± 8.3 μg·L-1) > Pb (3.3 ± 4.4 μg·L-1) > As (2.9 ± 2.3 μg·L-1) > Cr (1.8 ± 2.4 μg·L-1) ≫ Cd (0.3 ± 1.1 μg·L-1) > Hg (0.05 ± 0.12 μg·L-1). Wavelet analyses show that river flow regimes and flooding dominated the periodic variations in metal pollution. Bayesian network suggests that the hydrochemical factors (i.e., TOC, TP, TN, pH, and EC) chemically influenced the metal mobility between water and sediments. Furthermore, the source apportionment computed by the Bayesian multivariate receptor model shows that the given element contamination was typically attributed to the geogenic sources (17.5, 95% confidence interval: 13.1-17.6%), urban and industrial sources (22.1, 18.0-27.2%), arable soil erosion (24.2, 16.4-31.5%), and historical anthropogenic activities (35.2, 32.8-43.3%). The results provided herein reveal that both the hydrochemical influence on metal mobility and the chronic disturbance from anthropogenic activities caused the long-term variation in trace metal pollution.

A wide range of trace elements contained in solution and suspension in surface waters in Amur River basin, which are subject to different anthropogenic load, has been considered. They have been found to vary widely because of their natural conditions, the impact of anthropogenic factors near cities (Khabarovsk, Komsomol’sk-on-Amur), and transboundary transport from the territory of China. Dissolved Mo, Br, and I are proposed as tracers of such transport.

The Metallic Trace Elements (MTEs) occur in the waters and sediments of wadis with low concentrations. Their mobility is controlled by several physico-chemical parameters (Eh, pH and dissolved oxygen). To determine the effect of these factors on the concentration and mobility of metallic trace elements (Cd, Cr, Cu, Fe, Pb and Zn), bimonthly monitoring of physico-chemical parameters was conducted during the period from November 2013 to January 2015. The results obtained show that the MTEs are very influenced by the variations of the pH and Eh. Under natural conditions, neutral to basic pH and oxidizing medium, only lead is present in waters with high values, indicating its solubility in water and its release of sediments. Other MTEs have high concentrations in sediments, indicating their adsorption and/or chemical precipitation trapping qualities. The spatial monitoring of these MTEs shows the effect of the confluence of the wadi with its tributaries on the variation of its concentrations.

In this paper, the geographic distribution of arsenic and other trace elements in surface waters of 39 lotic ecosystems of central Argentina was evaluated. Manganese and arsenic were the most conspicuous elements, being present in 82% and 59% of the sampled ecosystems of this region, respectively. As concentration averaged 113.69 μg L⁻¹ varying between 55 and 198 μg L⁻¹, other trace elements were hardly detected or not detected at all. It was remarkable the absence of detectable concentrations of anthropogenically derived metals as lead (Pb) and cadmium (Cd).

Occurrence of trace elements in water, fish and shellfish from estuarine and coastal areas of the east coast of Bangladesh, northeast Bay of Bengal: Human health risk implications.

The Mahaweli is the largest river basin in Sri Lanka that provides water to the dry zone region through multipurpose irrigation schemes . Selenium, arsenic, cadmium, and other bioimportant trace elements in surface waters of the upper Mahaweli River were measured using ICP-MS. Trace element levels were then compared with water from two other rivers (Maha Oya, Kalu Ganga) and from six dry zone irrigation reservoirs. Results showed that the trace metal concentrations in the Mahaweli upper catchment were detected in the order of Fe>Cu>Zn>Se>Cr>Mn>As>Ni>Co>Mo. Remarkably high levels of Ca, Cr, Co, Ni, Cu, As, and Se were observed in the Mahaweli Basin compared to other study rivers. Considerably high levels of Cr, Mn, Fe, Co, Ni, Cu, Zn, As, and Se were found in upstream tributaries of the Mahaweli River. Such metals possibly originated from phosphate and organic fertilizers that are heavily applied for tea and vegetable cultivations within the drainage basin. Cadmium that is often attributed to the etiology of unknown chronic kidney diseases in certain parts of the dry zone is much lower than previously reported levels. Decrease in these metals in the lower part of the Mahaweli River could be due to adsorption of trace metals onto sediment and consequent deposition in reservoirs.

The way we learn has changed as a result of advances in science and technology and the growing influence of information technology on daily life. E-learning is a relatively new concept in educational settings, but it has progressively earned a spot in modern training techniques. Data mining methods are used in many applications, particularly in elearning. Comparing direct communications difficult to track the student involvement in online courses. To overcome the above mentioned issues the existing methods are analyzed various techniques of data mining and optimization methods which can be applied to bring out hidden knowledge from the educational data? In the e-learning work, preprocessing, feature selection and classification process are performed to improve the Kalboard 360 educational dataset accuracy significantly. This study focused on educational data mining (EDM) and employed decision trees, support vector machines (SVM), multilayer perceptrons, and the naive bayesian C4.5 algorithm among other classification techniques to assess the features for the prediction of students’ behaviour and academic achievement. This paper analyzes the advantages and shortcomings of each algorithm applied to data set on the basis of accuracy for various classification techniques like Multilayer Perceptron based ANN algorithm (MP-ANN), Decision Tree (DT) C4.5 and Hybrid Regression model with Multi Label Classification (HR-MLC) approach. The experimental result shows that the HR-MLC algorithm provides better performance in terms of higher accuracy, precision, recall and f-measure rather than the other existing methods.

The comparisons of data mining techniques for the predictive accuracy of probability of default of credit card clients

The aim of this research is to explore and analyze the benefits of risk classification methods by using data mining techniques on premium ratemaking in nonlife insurance. We rely on generalized linear models (GLMs) framework for nonlife premiums and examine the impact of specific data mining techniques on classifications of risk in motor hull insurance in Bosnia and Herzegovina. We study this relationship in an integrated framework considering a standard risk model based on the application of Poisson GLM for claims frequency estimate. Although GLM is a widely used method to determine insurance premiums, improvements of GLM by using the data mining methods identified in this paper may solve practical challenges for the risk models. The application of the data mining method in this paper aims to improve the results in the process of nonlife insurance premium ratemaking. The improvement is reflected in the choice of predictors or risk factors that have an impact on insurance premium rates. The following data mining methods for the selection of prediction variables were investigated: forward stepwise and neural networks. We provide strong and robust evidence that the use of data mining techniques influences premium ratemaking in nonlife insurance.KeywordsRatemakingForward stepwiseNeural networksNonlife insuranceGLM

In today's conditions, customer loyalty has gained importance with the increase in the competitive environment between companies, the development of marketing strategies and the improvement of companies. Therefore, it is essential to acquire customers for a company to survive. Retaining an existing customer in the telecommunication sector is less costly than gaining a new customer. Customer churn analysis is the process of predicting customers with high abandonment requests by examining the offers and utilizable behaviors. Customer churn analysis provides services to develop various campaigns aiming to increase the company’s loyalty by predicting the customers who are planning to move to another company. In this way, it gives the company a competitive advantage. This study aims to make predictions by developing models for customer churns through data mining and machine learning methods in the telecommunication sector. In addition, we believe that the application in this article will contribute to data analysts and academicians who will want to analyze customer churn with different data sets in telecommunication and other sectors in the future. The analysis in this study is carried out on a data set obtained from an open-access database, including 20 transaction records for the customer from 7043 customers and whether the customer left the company. Among the data mining methods, Random Forest (RF), Support Vector Machines (SVM) and Multilayer Artificial Neural Networks (ANN) are modeled in open-source Phyton environment. The results have shown that ANN has fared better at classifying customers than other machine learning methods.

Predictive analytics has emerged as a powerful tool in the field of financial engineering, enabling organizations to forecast trends, optimize strategies, and mitigate risks. Data mining techniques play a crucial role in extracting meaningful patterns from large volumes of financial datasets, contributing to the development of predictive modeling techniques. The purpose of this paper is to explore the application of various data mining techniques in predictive analytics within the financial sector, with a focus on their effectiveness in risk management, fraud detection, and portfolio optimization. Machine learning algorithms, such as decision trees, support vector machines, and neural networks, are increasingly being employed to model complex financial systems and enhance predictive accuracy. Additionally, techniques like clustering, association rule mining, and time series analysis are used to uncover hidden relationships and temporal patterns in financial datasets. Despite the significant advancements in these techniques, challenges remain in terms of data quality, computational complexity, and interpretability of results. This paper addresses these challenges and provides a comprehensive review of the most widely used data mining methods in financial engineering. Furthermore, it discusses the potential for future advancements in data mining algorithms and their integration with other emerging technologies like artificial intelligence and big data analytics. The findings indicate that while data mining offers significant potential in financial analytics, careful consideration must be given to the selection of appropriate techniques and the proper handling of data to ensure reliable predictions. Ultimately, this paper aims to contribute to the ongoing development of robust predictive modeling techniques in financial engineering by emphasizing the importance of data mining as an indispensable tool.

One of the most difficult tasks in the whole KDD process is to choose the right data mining technique, as the commercial software tools provide more and more possibilities together and the decision requires more and more expertise on the methodological point of view. Indeed, there are a lot of data mining techniques available for an environmental scientist wishing to discover some model from her/his data. This diversity can cause some troubles to the scientist who often have not a clear idea of what are the available methods, and moreover, use to have doubts about the most suitable method to be applied to solve a concrete domain problem. Within the data mining literature there is not a common terminology. A classification of the data mining methods would greatly simplify the understanding of the whole space of available methods. In this work, a classification of most common data mining methods is presented in a conceptual map which makes easier the selection process. Also an intelligent data mining assistant is presented. It is oriented to provide model/algorithm selection support, suggesting the user the most suitable data mining techniques for a given problem.