The Estimation of the Relationships between Water-Natural Gas Usage and Discharge-Emission Permission by Using Binary Logistic Model for the Industrial Establishments

The Thrace region is one of the most important agricultural and industrial regions in Turkey with its plain, deep and productive soil structure. The most important problem is industrial based water and air pollution in this region. In this study, 90 out of 541 establishments that use industrial water in the region were studied representing approximately 17% of the total population. The relationship between water-natural gas usage and discharge-emission permission is estimated by using the binary logistic model. In addition to the economic and social benefits, there are also economic and social losses due to environmental pollution because of the rapid industrialization in the Thrace Region. For example; the rice producers that use Ergene River for irrigation purposes lose efficiency and quality in various levels due to water pollution. According to the research results, an increase in wastewater is a factor that decreases the probability of obtaining discharge permission. Moreover, excessive use of underground water results in decreases in static and dynamic water levels in the region. Because of this result, it is more advantageous for sustainable water management to be directed to the sectors that do not use water for industrial purposes instead of the sectors that use water in their processes.     

A novel method as a diagnostic tool for the detection of influential observations in the Cox proportional hazards model

It is important that the process of studying and modelling the prognosis of disability should be conducted using time-to-event data, as the dynamic nature of disability could cause intervention on the modifiable (prognostic) factors, thus changing the course to a more favourable outcome. In disability research, the Cox PH model is frequently used to identify prognostic factors for the life expectancy of people with disabilities and to evaluate the treatment effects on the time to event. Accurate detection of influential observations is an important factor when fitting the Cox PH model, as influential observations in the Cox PH model can cause model misspecification, inaccurately determined factors, missed valuable biological information and/or violation of the proportional hazard assumption. In this paper, a novel multiple case detection method for influential observations is recommended in the Cox model. The aim of the paper is to inform clinicians and researchers who use the Cox PH model for describing the survival time as a function of multiple prognostic factors, regarding the importance of the detection of influential observations that can lead to misleading conclusions if they are present in the data set. The efficiency of the proposed method is presented through the real dataset. Additionally, in the specific case of North Cyprus, the aim is emphasize the importance of survival modelling studies that determine the prognostic factors affecting the lives of people with disabilities, to improve life quality and to develop a plan for healthier and higher quality life styles programmes for people with disabilities. As a first step, it is recommended that a system of database records of disabilities should be established and maintained by the government to raise public awareness.

     

采用神经网络技术研究色彩管理中数字打样对印刷质量的影响

本文的主要目的是采用人工神经网络（ANN）技术研究色彩管理中数字打样系统对印刷质量的影响。非涂布纸和涂布纸标准数据的R^2值分别为0．99702和0．99688,而测试数据的R^2值分别为0．994707和0．99629。这充分说明ANN方法能够准确地对色彩管理系统进行评价。测试输出结果表明,ANN模型可以可靠地评价数码打样系统中色彩管理对印刷质量的影响。