scholarly journals Residuals in the modelling of pollution concentration depending on meteorological conditions and traffic flow, employing decision trees

2018 ◽  
Vol 23 ◽  
pp. 00016 ◽  
Author(s):  
Joanna A. Kamińska

Two data mining methods – a random forest and boosted regression trees – were used to model values of roadside air pollution depending on meteorological conditions and traffic flow, using the example of data obtained in the city of Wrocław in the years 2015–2016. Eight explanatory variables – five continuous and three categorical – were considered in the models. A comparison was made of the quality of the fit of the models to empirical data. Commonly used goodness-of-fit measures did not imply a significant preference for either of the methods. Residual analysis was also performed; this showed boosted regression trees to be a more effective method for predicting typical values in the modelling of NO2, NOx and PM2.5, while the random forest method leads to smaller errors when predicting peaks.

Mnemosyne ◽  
2015 ◽  
Vol 68 (5) ◽  
pp. 794-813
Author(s):  
Cornelis van Tilburg

The founding of a city requires certain hygienic and meteorological conditions. The climate must be moderate, neither too hot, nor too cold; neither too dry, nor too moist; fresh air and water are crucial. Ancient medical writers such as the authors of the Hippocratic Corpus, Celsus and Galen prescribe ideal conditions for the city. Wind-directions, local climate (heat, cold, humidity), quantity and quality of air and water and a clean environment were crucial factors to establish a healthy city. Did their opinions correspond with the opinions of non-medical ancient sources like Vitruvius, Varro, and Columella? And, finally, were these conditions really realised in practice, as proved by excavations? According to his book Res rusticae, the Roman author M. Terentius Varro improved the hygienic situation by cleaning polluted air, when he changed the position of doors and windows. If this story is true, there is evidence that there was some knowledge of improving health, bringing theory into practice.


10.1068/c0427 ◽  
2005 ◽  
Vol 23 (3) ◽  
pp. 317-336 ◽  
Author(s):  
Mark Birkin ◽  
Graham Clarke ◽  
Phil Gibson ◽  
Roger Dewhurst ◽  
Jacqui Bobby

This paper is concerned with modelling variations in the use of health-care services between small geographic areas. A range of potential explanatory variables are identified from a review of previous literature, ranging from social, economic, and demographic factors through access to services, and practitioner characteristics, to new measures of behaviour and lifestyle. Real admissions data for the city of Leeds relating to a variety of services over a three-year period are introduced to calibrate a series of utilisation models. It is argued that the strength of the goodness of fit makes these models potentially useful in the evaluation of resource allocation between service providers. By providing better global models of usage it is possible to examine small-area outliers to highlight areas where revealed demand, or usage, is not reflecting need as much as it should. In particular, this paper demonstrates the importance of lifestyle preferences in modelling the utilisation of health-care services.


Author(s):  
Gabriel Alarcón ◽  
Jorge Díaz ◽  
Mauro Vela ◽  
Mishari García ◽  
José Gutiérrez

<p class="Default">El estudio cuantifica las áreas deforestadas en una de las zonas más impactadas por el hombre, al suroeste de la ciudad de Puerto Maldonado (Puerto Maldonado – Inambari), las imágenes Landsat TM 5 y 8 OLI fueron procesados utilizando una clasificación semiautomática denominada “Random Forest” propuesto por la <a href="http://reddcommunity.org/link/mesa-de-servicios-ambientales-y-redd-de-madre-de-dios">Mesa de Servicios Ambientales y Reducción de Emisiones de Gases de Efecto Invernadero Causadas por la Deforestación y Degradación de los Bosques de la Región Madre de Dios</a>, Perú. La metodología incluyo procedimientos de documentación, verificación y validación que permitieron evaluar la calidad de la información generada y los datos reportados. Los resultados de la investigación reportan; una superficie deforestada para los años 1999-2013 de 55426 ha, que representa una tasa anual de cambio de cobertura de – 0,22% y una deforestación anual promedio de 3 246 ha/año. De ello se deducen para los años 1999-2008, 2008-2011 y 2011-2013 tasas anuales de cambio de – 0,18%, -0,30% y -0,31%, con una deforestación anual promedio de 2 594 ha/año, 4 427 ha/año y 4 410 ha/año respectivamente. Los cambios producidos en el área de estudio, muestran la sistemática conversión de bosque a deforestación para el año 2008-2011 con 29 478 ha, siendo la mayor responsable el avance de la minería aurífera aluvial influenciada por la pavimentación de la vía interoceánica y por el alza del precio del oro como el agente principal de la deforestación, y en menor orden, la ampliación de la frontera agrícola, la ganadería y la actividad forestal.</p><p align="center"> </p><p align="center"><strong>ABSTRACT</strong></p><p>The study quantifies the deforested areas in one of the areas most affected by  humans, southwest in the city of Puerto Maldonado (Puerto Maldonado - Inambari), the Landsat images TM 5 and 8 OLI were processed using a semiautomatic classification called "Random Forest" proposed by Mesa de Servicios Ambientales y Reducción de Emisiones de Gases de Efecto Invernadero Causadas por la Deforestación y Degradación de los Bosques in the region of Madre de Dios, Peru. The methodology included documentation procedures, verification and validation to assess the quality of information generated and the data reported. The results of the investigation report; a deforested for years 1999-2013 of 55426 ha, which represents an annual rate of change of coverage of -0,22% and an average annual deforestation of 3246 ha/year. It deducted for the years 1999-2008, 2008-2011 and 2011-2013 annual rates of change of -0,18%, -0,30% and -0,31%, with an average annual deforestation of 2594 ha/year, 4427 ha/year and 4,410 ha/year respectively. Changes in the study area, show the systematic conversion of forest to deforestation for the year 2008-2011 with 29 478 ha, being most responsible advancing of alluvial gold mining influenced by the paving of the interoceanic highway and the gold price rises as the main agent of deforestation, and lower order, expanding the agricultural frontier, cattle rising and forestry.</p><p><strong><em>Key words: </em></strong>Cover change, Annual change rate, Random Forest, alluvial Gold Mining.</p><p> </p><p> </p><p><strong> </strong></p><p><strong> </strong></p><p><strong> </strong></p>


2017 ◽  
Vol 13 (1) ◽  
pp. 147 ◽  
Author(s):  
Juan Aznar ◽  
Josep M Sayeras ◽  
Alba Rocafort ◽  
Jorge Galiana

Purpose: The aim of this study is to analyze the existence of a relationship between the presence of nearby substitute products, mainly Airbnb flats or rooms, and the effect on the revenue and profitability of hotels.Design/methodology/approach: The empirical study is based on the analysis of financial information provided in the annual reports of a sample consisting of 43 hotels (11.78% of the population). As an explanatory variable for profitability, we have considered the presence of apartments listed in Airbnb that are no farther than one kilometer from each hotel. Considering that most of the variables used do not follow a normal distribution, the existence of a relationship between profitability and the explanatory variables has been tested using non-parametric tests, namely, the Spearman correlation coefficient and Kruskall-Wallis test.Findings: We found a positive correlation between presence of Airbnb apartments and return on equity. This fact can be explained by considering the presence of Airbnb apartments as a variable that measures the attractiveness of the location from a tourist’s point of view. Hotels located near the city center or main tourist areas of the city have a higher level of profitability. We also found no evidence of any relationship between profitability and star category; 4- and 3-star hotels have experienced, on average, a lower drop in revenues between 2008 and 2013 and they also obtained a higher average level of profitability in 2013 as compared to the upper segment of 5-star hotels.Research limitations/implications: This research has been conducted in the city of Barcelona. Future research using the same methodology should be applied to other cities with an important hospitality sector to reinforce our findings. The main implications of this research refer to the importance of location as a key strategic variable in hospitality, and to the change in the system customers use to evaluate the quality of a hotel, according to which the traditional star category system has been partially replaced by new sources of information available through the new communication technologies.Originality/value: This paper is one of the first contributions that scholars have made to obtain a deeper understanding of the effects of the new forms of the sharing economy, focusing on the hospitality industry.


2019 ◽  
Vol 17 (2) ◽  
pp. 175-181 ◽  
Author(s):  
Alexander Novikov ◽  
Ivan Novikov ◽  
Anastasia Shevtsova
Keyword(s):  

2021 ◽  
Vol 10 (12) ◽  
pp. 836
Author(s):  
Jiansheng Wu ◽  
Yun Qian ◽  
Yuan Wang ◽  
Na Wang

During the COVID-19 lockdown in Wuhan, transportation, industrial production and other human activities declined significantly, as did the NO2 concentration. In order to assess the relative contributions of different factors to reductions in air pollutants, we implemented sensitivity experiments by Random Forest (RF) models, with the comparison of the contributions of meteorological conditions, human mobility, and emissions from industry and households between different periods. In addition, we conducted scenario analyses to suggest an appropriate limit for control of human mobility. Different mechanisms for air pollutants were shown in the pre-pandemic, pre-lockdown, lockdown, and post-pandemic periods. Wind speed and the Within-city Migration index, representing intra-city mobility intensity, were excluded from stepwise multiple linear models in the pre-lockdown and lockdown periods. The results of sensitivity experiments show that, in the COVID-19 lockdown period, 73.3% of the reduction can be attributed to decreased human mobility. In the post-pandemic period, meteorological conditions control about 42.2% of the decrease, and emissions from industry and households control 40.0%, while human mobility only contributes 17.8%. The results of the scenario analysis suggest that the priority of restriction should be given to human mobility within the city than other kinds of human mobility. The reduction in the NO2 concentration tends to be smaller when human mobility within the city decreases by more than 70%. A limit of less than 40% on the control of the human mobility can achieve a better effect, especially in cities with severe traffic pollution.


2014 ◽  
Vol 4 (2) ◽  
pp. 101
Author(s):  
Ihsany Abdillah ◽  
Fitroh Adhilla

Market orientation is believed to be a source of competitive advantage that is difficult to imitate by competitors. This study aimed to determine the effect of market orientation on service quality, customer satisfaction, and customer loyalty Speedy PT Telkom in the city of Yogyakarta. The population in this study are all PT Telkom Speedy customers who live in city of Yogyakarta. Data analysis methods used in this study include descriptive and inferential methods, inferential method used is path analysis. The results of goodness of fit model shows that the X2 value, the value of CFI, and RMSEA values have had a good match. Based on the analysis of data shows that the determinants of customer loyalty Telkom Speedy Yogyakarta is a market orientation quality of service, and customer satisfaction.


2019 ◽  
Vol 44 (44) ◽  
pp. 93-101 ◽  
Author(s):  
Szymon Wójcik

AbstractIn this study, potential factors influencing the decisions made by citizens of the city of Łódź, Poland, regarding the choice of transportation mode used in their daily travel activities were examined. In addition to a brief literature review, an empirical study was performed. Data from a previous quality-of-life study were used to enhance the scope of explanatory variables in a regression model. In order to identify the determinants of travel behaviour, binary logistic regression models were used. The results show that socio-demographic characteristics of respondents and household access to a car most influenced transport mode choices. Also, the relationship between geographic distances and subjective opinions regarding public transport were found to be statistically significant. The determinants for choosing either public or private transportation varied.


Sign in / Sign up

Export Citation Format

Share Document