Wyniki wyszukiwania - BazTech

1

A Multivariate Technique to Develop Hybrid Water Quality Index of the Bengawan Solo River, Indonesia

Lusiana Evellin Dewi, Mahmudi Mohammad, Hutahaean Sarah Mega, Darmawan Arief, Buwono Nanik Retno, Arsad Sulastri, Musa Muhammad

Journal of Ecological Engineering

|

2022

|

Vol. 23, nr 2

123--131

EN

Surface water resource, such as river, is constantly contaminated by domestic and industrial pollutants. In order to properly manage the water resource, a composite index for water quality assessment, such as water quality index (WQI), has been designed to monitor and evaluate the properties of surface water. However, this index is quite subjective in terms of determination of relative weights. A principal component analysis (PCA) can be used to reduce the dimension and subjectivity of water quality variables. The purpose of this study was to implement the use of hybrid PCA and WQI methods to assess and monitor the water quality of the Bengawan Solo River, which is located in Java Island, Indonesia. The result suggested that COD, BOD, TSS, TDS, nitrate, nitrite, and ammonia were the main factors that determine water quality of the Bengawan Solo River. Furthermore, it was revealed that most samples from the river showed water quality status as slightly polluted. In addition to this, the seasonal variation of the PCWI values indicated a significant increase of water pollution in the Bengawan Solo River per year.

2

Applying latent class analysis in the identification of occupational accident patterns

Nowakowska Marzena, Pajęcki Michał

Zeszyty Naukowe. Organizacja i Zarządzanie / Politechnika Śląska

|

2020

|

z. 146

339--355

EN

Purpose: The objective of the study is to use selected data mining techniques to discover patterns of certain recurring mechanisms related to the occurrence of occupational accidents in relation to production processes. Design/methodology/approach: The latent class analysis (LCA) method was employed in the investigation. This statistical modeling technique enables discovering mutually exclusive homogenous classes of objects in a multivariate data set on the basis of observable qualitative variables, defining the class homogeneity in terms of probabilities. Due to a bilateral agreement, Statistics Poland provided individual record-level real data for the research. Then the data were preprocessed to enable the LCA model identification. Pilot studies were conducted in relation to occupational accidents registered in production plants in 2008-2017 in the Wielkopolskie voivodeship. Findings: Three severe accident patterns and two light accident patterns represented by latent classes were obtained. The classes were subjected to descriptive characteristics and labeling, using interpretable results presented in the form of probabilities classifying categories of observable variables, symptomatic for a given latent class. Research limitations/implications: The results from the pilot studies indicate the necessity to continue the research based on a larger data set along with the analysis development, particularly as regards selecting indicators for the latent class model characterization. Practical implications: The identification of occupational accident patterns related to the production process can play a vital role in the elaboration of efficient safety countermeasures that can help to improve the prevention and outcome mitigation of such accidents among workers. Social implications: Creating a safe work environment comprises the quality of life of workers, their families, thus affirming the enterprises' principles and values in the area of corporate social responsibility. Originality/value: The investigation showed that latent class analysis is a promising tool supporting the scientific research in discovering the patterns of occupational accidents. The proposed investigation approach indicates the importance for the research both in terms of the availability of non-aggregated occupational accident data as well as the type of value aggregation of the variables taken for the analysis.

3

Portfolio Inputs Selection from Imprecise Training Data

Raudys S., Raudys A., Biziuleviciene G., Pabarskaite Z.

Schedae Informaticae

|

2016

|

Vol. 25

177--188

EN

This paper explores very acute problem of portfolio secondary overfitting. We examined the financial portfolio inputs random selection optimization model and derived the equation to calculate the mean Sharpe ratio in dependence of the number of portfolio inputs, the sample size L used to estimate Sharpe ratios of each particular subset of inputs and the number of times the portfolio inputs were generated randomly. It was demonstrated that with the increase in portfolio complexity, and complexity of optimization procedure we can observe the over-fitting phenomena. Theoretically based conclusions were confirmed by experiments with artificial and real world 60,000-dimensional 12 years financial data.

4

Prediction of the result in race walking using regularized regression models

Przednowek K., Wiktorowicz K.

Journal of Theoretical and Applied Computer Science

|

2013

|

Vol. 7, nr 2

45--58

EN

The following paper presents the use of regularized linear models as tools to optimize training process. The models were calculated by using data collected from race-walkers’ training events. The models used predict the outcomes over a 3 km race and following a prescribed training plan. The material included a total of 122 training patterns made by 21 players. The methods of analysis include: classical model of OLS regression, ridge regression, LASSO regression and elastic net regression. In order to compare and choose the best method a cross-validation of the leave-one-out was used. All models were calculated using R language with additional packages. The best model was determined by the LASSO method which generates an error of about 26 seconds. The methodhas simplified the structure of the model by eliminating 5 out of 18 predictors.