Pub Date : 2023-06-13DOI: 10.59170/stattrans-2023-029
G. Kalton
At the beginning of the 20th century, there was an active debate about random selection of units versus purposive selection of groups of units for survey samples. Neyman’s (1934) paper tilted the balance strongly towards varieties of probability sampling combined with design-based inference, and most national statistical offices have adopted this method for their major surveys. However, nonprobability sampling has remained in widespread use in many areas of application, and over time there have been challenges to the Neyman paradigm. In recent years, the balance has tilted towards greater use of nonprobability sampling for several reasons, including: the growing imperfections and costs in applying probability sample designs; the emergence of the internet and other sources for obtaining survey data from very large samples at low cost and at high speed; and the current ability to apply advanced methods for calibrating nonprobability samples to conform to external population controls. This paper presents an overview of the history of the use of probability and nonprobability sampling from the birth of survey sampling at the time of A. N. Kiær (1895) to the present day.
{"title":"Probability vs. Nonprobability Sampling: From the Birth of Survey Sampling to the\u0000 Present Day","authors":"G. Kalton","doi":"10.59170/stattrans-2023-029","DOIUrl":"https://doi.org/10.59170/stattrans-2023-029","url":null,"abstract":"At the beginning of the 20th century, there was an active debate about random\u0000 selection of units versus purposive selection of groups of units for survey samples.\u0000 Neyman’s (1934) paper tilted the balance strongly towards varieties of probability\u0000 sampling combined with design-based inference, and most national statistical offices\u0000 have adopted this method for their major surveys. However, nonprobability sampling has\u0000 remained in widespread use in many areas of application, and over time there have been\u0000 challenges to the Neyman paradigm. In recent years, the balance has tilted towards\u0000 greater use of nonprobability sampling for several reasons, including: the growing\u0000 imperfections and costs in applying probability sample designs; the emergence of the\u0000 internet and other sources for obtaining survey data from very large samples at low cost\u0000 and at high speed; and the current ability to apply advanced methods for calibrating\u0000 nonprobability samples to conform to external population controls. This paper presents\u0000 an overview of the history of the use of probability and nonprobability sampling from\u0000 the birth of survey sampling at the time of A. N. Kiær (1895) to the present\u0000 day.","PeriodicalId":37985,"journal":{"name":"Statistics in Transition","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2023-06-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"43980508","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-06-13DOI: 10.59170/stattrans-2023-043
Anand Sharma, Vipin Sharma, Shekhar Tokas
This study empirically investigates the relationship between economic freedom and financial development in EU countries. Using panel data covering the years 2000–2017 and employing fixed effects, random effects, and the generalised method of moments (GMM), the paper examines the effect of economic freedom on financial development. The research results demonstrate that greater economic freedom is conducive to financial development in the EU. These findings remain robust to the use of an alternative index of economic freedom. The results imply that policies which promote economic freedom are likely to raise the level of a country’s financial development.
{"title":"Does economic freedom promote financial development? Evidence from EU\u0000 countries","authors":"Anand Sharma, Vipin Sharma, Shekhar Tokas","doi":"10.59170/stattrans-2023-043","DOIUrl":"https://doi.org/10.59170/stattrans-2023-043","url":null,"abstract":"This study empirically investigates the relationship between economic freedom and\u0000 financial development in EU countries. Using panel data covering the years 2000–2017 and\u0000 employing fixed effects, random effects, and the generalised method of moments (GMM),\u0000 the paper examines the effect of economic freedom on financial development. The research\u0000 results demonstrate that greater economic freedom is conducive to financial development\u0000 in the EU. These findings remain robust to the use of an alternative index of economic\u0000 freedom. The results imply that policies which promote economic freedom are likely to\u0000 raise the level of a country’s financial development.","PeriodicalId":37985,"journal":{"name":"Statistics in Transition","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2023-06-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"41390763","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-06-13DOI: 10.59170/stattrans-2023-031
R. Lehtonen
I would like to congratulate Professor Graham Kalton for his significant and inspiring article entitled as "Probability vs. Nonprobability Sampling: From the Birth of Survey Sampling to the Present Day". The article provides an elegant overview of the history of survey sampling, covering the purposive approaches that dominated the sampling field in the early days but from the 1940s, at least in official statistics, were gradually replaced entirely by probability-based approaches. Today we may be facing a paradigm shift again, but the direction is the opposite. Non-probability-based approaches are becoming viable, if not the only option, in fields that are moving towards big data and other new data sources and new methodological approaches.
{"title":"Comments on „Probability vs. Nonprobability Sampling: From the Birth of Survey\u0000 Sampling to the Present Day” by Graham Kalton","authors":"R. Lehtonen","doi":"10.59170/stattrans-2023-031","DOIUrl":"https://doi.org/10.59170/stattrans-2023-031","url":null,"abstract":"I would like to congratulate Professor Graham Kalton for his significant and\u0000 inspiring article entitled as \"Probability vs. Nonprobability Sampling: From the\u0000 Birth of Survey Sampling to the Present Day\". The article provides an elegant\u0000 overview of the history of survey sampling, covering the purposive approaches that\u0000 dominated the sampling field in the early days but from the 1940s, at least in official\u0000 statistics, were gradually replaced entirely by probability-based approaches. Today we\u0000 may be facing a paradigm shift again, but the direction is the opposite.\u0000 Non-probability-based approaches are becoming viable, if not the only option, in fields\u0000 that are moving towards big data and other new data sources and new methodological\u0000 approaches.","PeriodicalId":37985,"journal":{"name":"Statistics in Transition","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2023-06-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"43648704","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-06-13DOI: 10.59170/stattrans-2023-032
J. Gershunskaya, P. Lahiri
In this excellent overview of the history of probability and nonprobability sampling from the end of the nineteenth century to the present day, Professor Graham Kalton outlines the essence of past endeavors that helped to define philosophical approaches and stimulate the development of survey sampling methodologies. From the beginning, there was an understanding that a sample should, in some ways, resemble the population under study. In Kiær’s ideas of “representative sampling” and Neyman’s invention of probability-based approach, the prime concern of survey sampling has been to properly plan for representing characteristics of the finite population. Poststratification and other calibration methods were developed for the same important goal of better representation.
Graham Kalton教授在这篇关于十九世纪末至今概率和非概率抽样历史的精彩综述中，概述了过去努力的本质，这些努力有助于定义哲学方法并促进调查抽样方法的发展。从一开始，人们就认为样本在某些方面应该与所研究的人群相似。在Kiær的“代表性抽样”思想和Neyman发明的基于概率的方法中，调查抽样的主要关注点是正确规划代表有限总体的特征。后分层和其他校准方法是为了更好地表现同样的重要目标而开发的。
{"title":"Discussion of “Probability vs. Nonprobability Sampling: From the Birth of Survey\u0000 Sampling to the Present Day” by Graham Kalton","authors":"J. Gershunskaya, P. Lahiri","doi":"10.59170/stattrans-2023-032","DOIUrl":"https://doi.org/10.59170/stattrans-2023-032","url":null,"abstract":"In this excellent overview of the history of probability and nonprobability\u0000 sampling from the end of the nineteenth century to the present day, Professor Graham\u0000 Kalton outlines the essence of past endeavors that helped to define philosophical\u0000 approaches and stimulate the development of survey sampling methodologies. From the\u0000 beginning, there was an understanding that a sample should, in some ways, resemble the\u0000 population under study. In Kiær’s ideas of “representative sampling” and Neyman’s\u0000 invention of probability-based approach, the prime concern of survey sampling has been\u0000 to properly plan for representing characteristics of the finite population.\u0000 Poststratification and other calibration methods were developed for the same important\u0000 goal of better representation.","PeriodicalId":37985,"journal":{"name":"Statistics in Transition","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2023-06-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"44689830","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-06-13DOI: 10.59170/stattrans-2023-038
Kuldeep Kumar Tiwari, Vishwantra Sharma
In real-world surveys, non-response and measurement errors are common, therefore studying them together seems rational. Some population mean estimators are modified and studied in the presence of non-response and measurement errors. Bias and mean squared error expressions are derived under different cases. For all estimators, a theoretical comparison is made with the sample mean per unit estimator. The Monte-Carlo simulation is used to present a detailed picture of all estimators’ performance.
{"title":"Efficient estimation of population mean in the presence of non-response and\u0000 measurement error","authors":"Kuldeep Kumar Tiwari, Vishwantra Sharma","doi":"10.59170/stattrans-2023-038","DOIUrl":"https://doi.org/10.59170/stattrans-2023-038","url":null,"abstract":"In real-world surveys, non-response and measurement errors are common, therefore\u0000 studying them together seems rational. Some population mean estimators are modified and\u0000 studied in the presence of non-response and measurement errors. Bias and mean squared\u0000 error expressions are derived under different cases. For all estimators, a theoretical\u0000 comparison is made with the sample mean per unit estimator. The Monte-Carlo simulation\u0000 is used to present a detailed picture of all estimators’ performance.","PeriodicalId":37985,"journal":{"name":"Statistics in Transition","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2023-06-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"48147300","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-06-13DOI: 10.59170/stattrans-2023-045
S. Chugaievska, G. Dehnel, Andrey Targonskii
The latest National Population Census in Poland, like in many EU countries, was conducted in 2021. In Ukraine, during the entire period of independence, a national census was conducted only once, in 2001, while the following rounds kept being postponed. In 2019, a trial census was conducted in several regions of central Ukraine. The working hypothesis is that for the organisation and conduct of the All-Ukrainian Population Census in Ukraine, it is important to use the experience of EU countries in this field (and Poland’s experience in particular). The purpose of the article is to substantiate the theoretical foundations and comparative analysis of the processes of conducting censuses in Poland and Ukraine and to study the factors of increasing the level of respondent participation. The article reviews a number of socio-economic factors in the light of the comparison of data census organisation in Poland and Ukraine. Several international indicators were investigated that could have an impact on respondent participation in statistical surveys and censuses. By applying factor analysis, it was possible to identify the factors that could increase the level of respondent participation. To solve these demographic challenges, the following activity should be considered: financial support for the successful functioning of economic entities, improvement of social infrastructure, friendly environment for innovation and investment, and peace and political stability in Ukraine.
{"title":"Census administration in Ukraine: insight into the Polish experience in the context\u0000 of international indicators analysis","authors":"S. Chugaievska, G. Dehnel, Andrey Targonskii","doi":"10.59170/stattrans-2023-045","DOIUrl":"https://doi.org/10.59170/stattrans-2023-045","url":null,"abstract":"The latest National Population Census in Poland, like in many EU countries, was\u0000 conducted in 2021. In Ukraine, during the entire period of independence, a national\u0000 census was conducted only once, in 2001, while the following rounds kept being\u0000 postponed. In 2019, a trial census was conducted in several regions of central Ukraine.\u0000 The working hypothesis is that for the organisation and conduct of the All-Ukrainian\u0000 Population Census in Ukraine, it is important to use the experience of EU countries in\u0000 this field (and Poland’s experience in particular). The purpose of the article is to\u0000 substantiate the theoretical foundations and comparative analysis of the processes of\u0000 conducting censuses in Poland and Ukraine and to study the factors of increasing the\u0000 level of respondent participation. The article reviews a number of socio-economic\u0000 factors in the light of the comparison of data census organisation in Poland and\u0000 Ukraine. Several international indicators were investigated that could have an impact on\u0000 respondent participation in statistical surveys and censuses. By applying factor\u0000 analysis, it was possible to identify the factors that could increase the level of\u0000 respondent participation. To solve these demographic challenges, the following activity\u0000 should be considered: financial support for the successful functioning of economic\u0000 entities, improvement of social infrastructure, friendly environment for innovation and\u0000 investment, and peace and political stability in Ukraine.","PeriodicalId":37985,"journal":{"name":"Statistics in Transition","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2023-06-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"48551552","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-03-15DOI: 10.59170/stattrans-2023-028
L. Campanelli
Using the Euclidean distance statistical test of Benford’s law, we analyse the COVID-19 weekly case counts by country. While 62% of the 100 countries and territories considered in the present study conforms to Benford’s law at a significant level of α = 0.05 and 17% at a significant level of 0.01 ≤ α < 0.05, the remaining 21% shows a deviation from it (p values smaller than 0.01). In particular, 5% of the countries ‘break’ Benford’s law with a p value smaller than 0.001.
{"title":"Breaking Benford’s law: a statistical analysis of COVID-19 data using the Euclidean\u0000 distance statistic","authors":"L. Campanelli","doi":"10.59170/stattrans-2023-028","DOIUrl":"https://doi.org/10.59170/stattrans-2023-028","url":null,"abstract":"Using the Euclidean distance statistical test of Benford’s law, we analyse the\u0000 COVID-19 weekly case counts by country. While 62% of the 100 countries and territories\u0000 considered in the present study conforms to Benford’s law at a significant level of α =\u0000 0.05 and 17% at a significant level of 0.01 ≤ α < 0.05, the remaining 21% shows a\u0000 deviation from it (p values smaller than 0.01). In particular, 5% of the countries\u0000 ‘break’ Benford’s law with a p value smaller than 0.001.","PeriodicalId":37985,"journal":{"name":"Statistics in Transition","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2023-03-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"45062896","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-03-15DOI: 10.59170/stattrans-2023-026
E. Ibtissam, Mohamed Hadini, Adil Chebir, Ben Ali Mohamed, Echchelh Adil
Implemented by an increasing number of organisations worldwide, the ISO 9001 standard for quality management received considerable attention in the existing literature. Researchers worldwide have found positive, negative and even mixed effects of ISO 9001 certification on firms’ performance, while in Morocco this issue has been rarely examined. It is the combination of these observations that led to this study. The aim of this paper is to test and validate a causal model designed to measure the performance of an ISO 9001 certified Quality Management System (QMS) and its impact on a company’s financial performance. By means of this causal analysis/model, the study examines the relationship between: • QMS and the financial performance of 41 companies based in Morocco; • the management responsibility process and all the QMS processes; • the management resources process and all the QMS processes; • the organisational and financial performance of the studied companies. All of the considered firms are part of the service industry and range from medium-sized to large companies. The data gathered in this study have been instrumental in devising actionable insights. The Statistical Package for the Social Sciences (SPSS) was the statistical software platform that enabled the use of a linear regression analysis to prove the positive correlation between the above-mentioned elements.
ISO 9001质量管理标准由世界各地越来越多的组织实施，在现有文献中受到了相当大的关注。世界各地的研究人员发现，ISO9001认证对企业绩效的影响有积极的、消极的，甚至是混合的，而在摩洛哥，这一问题很少得到研究。正是这些观察结果的结合导致了这项研究。本文的目的是测试和验证一个因果模型，该模型旨在衡量ISO9001认证的质量管理体系（QMS）的绩效及其对公司财务绩效的影响。通过这种因果分析/模型，本研究考察了以下两者之间的关系：•质量管理体系与41家摩洛哥公司的财务业绩；•管理责任流程和所有QMS流程；•管理资源流程和所有QMS流程；•被研究公司的组织和财务表现。所有被考虑的公司都是服务业的一部分，从中型到大型公司都有。在这项研究中收集的数据有助于设计可行的见解。社会科学统计软件包（SPSS）是一个统计软件平台，能够使用线性回归分析来证明上述因素之间的正相关性。
{"title":"Proposal of a causal model measuring the impact of an ISO 9001 certified Quality\u0000 Management System on financial performance of Moroccan service-based companies","authors":"E. Ibtissam, Mohamed Hadini, Adil Chebir, Ben Ali Mohamed, Echchelh Adil","doi":"10.59170/stattrans-2023-026","DOIUrl":"https://doi.org/10.59170/stattrans-2023-026","url":null,"abstract":"Implemented by an increasing number of organisations worldwide, the ISO 9001\u0000 standard for quality management received considerable attention in the existing\u0000 literature. Researchers worldwide have found positive, negative and even mixed effects\u0000 of ISO 9001 certification on firms’ performance, while in Morocco this issue has been\u0000 rarely examined. It is the combination of these observations that led to this study. The\u0000 aim of this paper is to test and validate a causal model designed to measure the\u0000 performance of an ISO 9001 certified Quality Management System (QMS) and its impact on a\u0000 company’s financial performance. By means of this causal analysis/model, the study\u0000 examines the relationship between: • QMS and the financial performance of 41 companies\u0000 based in Morocco; • the management responsibility process and all the QMS processes; •\u0000 the management resources process and all the QMS processes; • the organisational and\u0000 financial performance of the studied companies. All of the considered firms are part of\u0000 the service industry and range from medium-sized to large companies. The data gathered\u0000 in this study have been instrumental in devising actionable insights. The Statistical\u0000 Package for the Social Sciences (SPSS) was the statistical software platform that\u0000 enabled the use of a linear regression analysis to prove the positive correlation\u0000 between the above-mentioned elements.","PeriodicalId":37985,"journal":{"name":"Statistics in Transition","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2023-03-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"44845295","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-03-15DOI: 10.59170/stattrans-2023-020
Adam Chwila
The COVID-19 pandemic has had a huge impact both on the global economy and on everyday life in all countries all over the world. In this paper, we propose several possible machine learning approaches to forecasting new confirmed COVID-19 cases, including the LASSO regression, Gradient Boosted (GB) regression trees, Support Vector Regression (SVR), and Long-Short Term Memory (LSTM) neural network. The above methods are applied in two variants: to the data prepared for the whole Poland and to the data prepared separately for each of the 16 voivodeships (NUTS 2 regions). The learning of all the models has been performed in two variants: with the 5-fold time-series cross-validation as well as with the split into the single train and test subsets. The computations in the study used official statistics from government reports from the period of April 2020 to March 2022. We propose a setup of 16 scenarios of the model selection to detect the model characterized by the best ex-post prediction accuracy. The scenarios differ from each other by the following features: the machine learning model, the method for the hyperparameters selection and the data setup. The most accurate scenario for the LASSO and SVR machine learning approaches is the single train/test dataset split with data for the whole Poland, while in case of the LSTM and GB trees it is the cross validation with data for whole Poland. Among the best scenarios for each model, the most accurate ex-post RMSE is obtained for the SVR. For the model performing best in terms of the ex-post RMSE, the interpretation of the outcome is conducted with the Shapley values. The Shapley values make it possible to present the impact of auxiliary variables in the machine learning model on the actual predicted value. The knowledge regarding factors that have the strongest impact on the number of new infections can help companies to plan their economic activity during turbulent times of pandemics. We propose to identify and compare the most important variables that affect both the train and test datasets of the model.
{"title":"The prediction of new Covid-19 cases in Poland with machine learning models","authors":"Adam Chwila","doi":"10.59170/stattrans-2023-020","DOIUrl":"https://doi.org/10.59170/stattrans-2023-020","url":null,"abstract":"The COVID-19 pandemic has had a huge impact both on the global economy and on\u0000 everyday life in all countries all over the world. In this paper, we propose several\u0000 possible machine learning approaches to forecasting new confirmed COVID-19 cases,\u0000 including the LASSO regression, Gradient Boosted (GB) regression trees, Support Vector\u0000 Regression (SVR), and Long-Short Term Memory (LSTM) neural network. The above methods\u0000 are applied in two variants: to the data prepared for the whole Poland and to the data\u0000 prepared separately for each of the 16 voivodeships (NUTS 2 regions). The learning of\u0000 all the models has been performed in two variants: with the 5-fold time-series\u0000 cross-validation as well as with the split into the single train and test subsets. The\u0000 computations in the study used official statistics from government reports from the\u0000 period of April 2020 to March 2022. We propose a setup of 16 scenarios of the model\u0000 selection to detect the model characterized by the best ex-post prediction accuracy. The\u0000 scenarios differ from each other by the following features: the machine learning model,\u0000 the method for the hyperparameters selection and the data setup. The most accurate\u0000 scenario for the LASSO and SVR machine learning approaches is the single train/test\u0000 dataset split with data for the whole Poland, while in case of the LSTM and GB trees it\u0000 is the cross validation with data for whole Poland. Among the best scenarios for each\u0000 model, the most accurate ex-post RMSE is obtained for the SVR. For the model performing\u0000 best in terms of the ex-post RMSE, the interpretation of the outcome is conducted with\u0000 the Shapley values. The Shapley values make it possible to present the impact of\u0000 auxiliary variables in the machine learning model on the actual predicted value. The\u0000 knowledge regarding factors that have the strongest impact on the number of new\u0000 infections can help companies to plan their economic activity during turbulent times of\u0000 pandemics. We propose to identify and compare the most important variables that affect\u0000 both the train and test datasets of the model.","PeriodicalId":37985,"journal":{"name":"Statistics in Transition","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2023-03-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"46593861","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-03-15DOI: 10.59170/stattrans-2023-017
İpek DEVECİ KOCAKOÇ, İstem KÖYMEN KESER
The coefficient of the variation function is a useful descriptive statistic, especially when comparing the variability of more than two curve groups, even when they have significantly different mean curves. Since the coefficient of variation function is the ratio of the mean and standard deviation functions, its particular property is that it shows the acceleration more explicitly than the standard deviation function. The aim of the study is twofold: to show that the functional coefficient of variation is more sensitive to abrupt changes than the functional standard deviation and to propose the utilisation of the functional coefficient of variation as an outlier detection tool. Several simulation trials have shown that the coefficient of the variation function allows the effects of outliers to be seen explicitly.
{"title":"Outlier detection based on the functional coefficient of variation","authors":"İpek DEVECİ KOCAKOÇ, İstem KÖYMEN KESER","doi":"10.59170/stattrans-2023-017","DOIUrl":"https://doi.org/10.59170/stattrans-2023-017","url":null,"abstract":"The coefficient of the variation function is a useful descriptive statistic,\u0000 especially when comparing the variability of more than two curve groups, even when they\u0000 have significantly different mean curves. Since the coefficient of variation function is\u0000 the ratio of the mean and standard deviation functions, its particular property is that\u0000 it shows the acceleration more explicitly than the standard deviation function. The aim\u0000 of the study is twofold: to show that the functional coefficient of variation is more\u0000 sensitive to abrupt changes than the functional standard deviation and to propose the\u0000 utilisation of the functional coefficient of variation as an outlier detection tool.\u0000 Several simulation trials have shown that the coefficient of the variation function\u0000 allows the effects of outliers to be seen explicitly.","PeriodicalId":37985,"journal":{"name":"Statistics in Transition","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2023-03-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"45531666","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：