Multinomial Logistic Regression Techniques Research Articles

Characterizing diseases using genetic and clinical variables: A data analytics approach

AbstractPredictive analytics is crucial in precision medicine for personalized patient care. To aid in precision medicine, this study identifies a subset of genetic and clinical variables that can serve as predictors for classifying diseased tissues/disease types. To achieve this, experiments were performed on diseased tissues obtained from the L1000 dataset to assess differences in the functionality and predictive capabilities of genetic and clinical variables. In this study, the k‐means technique was used for clustering the diseased tissue types, and the multinomial logistic regression (MLR) technique was applied for classifying the diseased tissue types. Dimensionality reduction techniques including principal component analysis and Boruta are used extensively to reduce the dimensionality of genetic and clinical variables. The results showed that landmark genes performed slightly better in clustering diseased tissue types compared to any random set of 978 non‐landmark genes, and the difference is statistically significant. Furthermore, it was evident that both clinical and genetic variables were important in predicting the diseased tissue types. The top three clinical predictors for predicting diseased tissue types were identified as morphology, gender, and age of diagnosis. Additionally, this study explored the possibility of using the latent representations of the clusters of landmark and non‐landmark genes as predictors for an MLR classifier. The classification models built using MLR revealed that landmark genes can serve as a subset of genetic variables and/or as a proxy for clinical variables. This study concludes that combining predictive analytics with dimensionality reduction effectively identifies key predictors in precision medicine, enhancing diagnostic accuracy.

Patterns and Determinants of Households' Energy Choices for Cooking as Well as Their Implications for Poverty Eradication in Nigeria

ABSTRACT: Adequate access to environmentally-friendly energy sources is crucial for addressing the problems of environmental degradation and adverse climate change as well as other developmental challenges. The literature on the patterns and determinants of households' energy choices for cooking is very scanty. Essentially, the scanty literature, in general, did not adequately cater for household composition and economies of scale in household consumption as well as give adequate consideration to the impact of ownership of education/entertainment appliances and means of communication by households on households' energy choices for cooking. It is very important to utilize adult equivalence scales and estimate of economies of scales in household consumption in the analysis of household consumption in order to make such analysis highly robust. It is also very important to include ownership of education/entertainment appliances and means of communication in the analysis of households' energy choices for cooking because they (the appliances and means of communication) are expected to provide adequate awareness on the harmful effects of dirty energy sources and such awareness can influence households' energy choices. Furthermore, none of the studies in the scanty literature discussed the major implications for poverty eradication of the patterns and determinants of households' energy choices for cooking. This study adequately addressed the highlighted research gaps. The study analyzed the patterns and determinants of households' energy choices for cooking in Nigeria and discussed their implications for poverty eradication in the country using, among other things, descriptive statistics and multinomial logistic regression technique. The study was based on a nationally representative survey data, obtained from the National Bureau of Statistics. The study found that the patterns of households' energy choices for cooking in Nigeria exhibit the operation of energy ladder hypothesis. The study also found that increased educational level of household head, ownership of electric food preservation appliance by household and satisfactory housing condition are among the factors that increase the likelihoods of the uses of transition fuels and very clean energy sources as well as reduce the likelihood of the use of traditional biomass fuels. Adequate measures should therefore be put in place to sufficiently increase households' incomes and guarantee optimal access to environmentally-friendly energy sources for cooking in the country; this will pave the way for poverty eradication in the country.

Supervised smoking facility access, harm reduction practices, and substance use changes during the COVID-19 pandemic: a community-engaged cross-sectional study

BackgroundThe potential public health benefits of supervised smoking facilities (SSFs) are considerable, and yet implementation of SSFs in North America has been slow. We conducted this study to respond to significant knowledge gaps surrounding SSF utilization and to characterize substance use, harm reduction practices, and service utilization following the onset of the COVID-19 pandemic.MethodsA questionnaire was self-administered at a single site by 175 clients using an outdoor SSF in Vancouver, Canada, between October–December 2020. Questionnaire responses were summarized using descriptive statistics. Multinomial logistic regression techniques were used to examine factors associated with increased SSF utilization.ResultsAlmost all respondents reported daily substance use (93% daily use of opioids; 74% stimulants). Most used opioids (85%) and/or methamphetamine (66%) on the day of their visit to the SSF. Respondents reported drug use practice changes at the onset of COVID-19 to reduce harm, including using supervised consumption sites, not sharing equipment, accessing medically prescribed alternatives, cleaning supplies and surfaces, and stocking up on harm reduction supplies. Importantly, 45% of SSF clients reported using the SSF more often since the start of COVID-19 with 65.2% reporting daily use of the site. Increased substance use was associated with increased use of the SSF, after controlling for covariates.ConclusionsClients of the SSF reported increasing not only their substance use, but also their SSF utilization and harm reduction practices following the onset of COVID-19. Increased scope and scale of SSF services to meet these needs are necessary.

Outage-induced power backup choice in Pakistan

Like in many other developing countries, power outages are prevalent in Pakistan. Consumers are compelled to explore alternative energy strategies when the overhead of the outage surpasses that of implementing a backup plan. Inverters, small generators, and solar energy systems (SESs) are among the common backup options. This study utilizes primary data collected from the twin cities of Islamabad and Rawalpindi to delve into the factors influencing consumers' perception-based choices regarding backup solutions. To account for various backup choices, we employ multinomial logistic regression techniques on a sample of 952 households. Our analysis reveals that perception-based variables, such as electricity theft and decreasing satisfaction, negatively impact the likelihood of selecting off-grid renewable options like solar technology. Moreover, we observe that consumers residing in rented houses prefer inverters and generators, while homeownership significantly influences the adoption of all three backup choices. Interestingly, the rural-urban location of the household does not influence the decision to opt for a SES. These findings carry important policy implications for energy-deficient regions. The transition towards solar energy is driven by the household's strategy to ensure uninterrupted energy backup while simultaneously reducing the cost of electricity consumption.

The influence of COVID-19 pandemic and coping strategies on work operation of nonfarm household enterprises in Nigeria.

The purpose of this study is to investigate how COVID-19 pandemic including some coping strategies such as hand wash with soap and food consumption influences work operation or performance of nonfarm household enterprises (NHEs) in Nigeria using 2020 Living Standard Measurement Survey data of 1728 sample size. This study departs from existing study in two ways: first, the study employs multinomial logistic regression technique to ascertain the determinants of work performance of nonfarm household enterprises in Nigeria. Second, the study focuses on nonfarm enterprises such as petty trade, road side automobile and cab drivers. The results show that COVID-19 pandemic is significant with negative influence on the work operation of NHEs in Nigeria. The result of the study also reveals that coping strategy such as hand wash with soap during the pandemic is an important driver of work performance or operation of NHEs in Nigeria. Another coping strategy like food consumption by nonfarm household enterprises shows insignificant influence on work operation which implies that there is no relationship between food consumption and work operation by NHEs in Nigeria. The policy recommendation of this study, among others, is that policies should focus on procurement of sanitary material for public use. This can be achieved through public sensitisation in terms of organising workshops and conferences.

Government social protection and households’ welfare during the Covid-19 pandemic in South Africa

PurposeThis study evaluates the impact of government social protection interventions on households’ welfare in South Africa.Design/methodology/approachThe study uses survey data comprising 393 observations and the multinomial logistic regression technique to analyse the effect of government interventions on households’ welfare. For robustness purposes, a negative binomial regression model is also estimated whose results corroborate the main results from the multinomial regression model.FindingsThe study’s findings show that government economic interventions through social protection significantly reduce the likelihood of a decrease in household income or consumption. COVID-19 grant/social relief of distress grant, unemployment insurance, tax relief and job protection and creation are all significant in sustaining household income and consumption.Practical implicationsThe findings have policy implications for social development. Specifically, the findings support the use of government social protection as a safety net for low-income groups in South Africa.Originality/valueThe study presents preliminary evidence on the effectiveness of several measures used to ameliorate the COVID-19-induced recession within the South African context.

The 30-year cardiovascular risk trajectories and their independently associated factors in participants of a Brazilian cohort (CUME Study).

We aimed to analyze the different trajectories of 30-year cardiovascular risk (CVR) and its independently associated factors in participants of the CUME Study, a prospective study with alumni from federal universities of Minas Gerais State, Brazil. In this study, 1,286 participants who answered the baseline (2016) and follow-up (2018 and 2020) questionnaires were included. Trajectories of CVR, according to the Framingham score, were identified with the latent class growth modelling technique with the use of the censored normal model. Analysis of the factors independently associated with each of the trajectories was conducted with multinomial logistic regression technique. Three CVR trajectories were identified: Low-Low (68.3%), Medium-Medium (26.2%), and High-High (5.5%). Male sex, living in a stable union, and having moderate and high intakes of ultra-processed foods were positively associated with the Medium-Medium and High-High CVR trajectories. Having non-healthcare professional training and working were positively associated with the Medium-Medium CVR trajectory, whereas being physically active was negatively associated with the High-High CVR trajectory. In conclusion, more than one-third of participants had CVR trajectories in the Medium-Medium and High-High categories. Food consumption and physical activity are modifiable factors that were associated with these trajectories; thus, implementing health promotion measures could help prevent the persistence or worsen of CVR. On the other hand, sociodemographic and labor characteristics are non-modifiable factors that were associated with Medium-Medium and High-High trajectories, which could help identify people who should be monitored with more caution by health services.

Air Quality prediction using Multinomial Logistic Regression

Nowadays, Artificial Intelligence (AI) plays a primary role in different applications like medicine, science, health, and finance. In the past five decades, the development and progress of technology have allowed artificial intelligence to take an essential role in human life. Air quality classification is an excellent example of this role. The use of AI in this domain allows humans to predict whether the air is polluted or not. In effect, monitoring air quality and providing periodic and direct statistics are essential requirements to ensure good air quality for individuals in the community. For this reason, a decision-making system is built to decide whether the air is clean or not. Based on this system's decision, necessary practices and measures are taken to improve air quality and ensure air sustainability. In this paper, the multinomial logistic regression technique is used to detect the air pollution level. The proposed method is applied to a real dataset that consists of 145 responses recorded from an air quality multi-sensor device containing chemical sensors. The used device was placed in New York City, USA, from 1/1/2021 to 7/1/2021 (one week) and is freely available for air quality sensors deployed in the field. The result shows the efficacity of this method in air pollution prediction.

Do risk attitude and trust moderate bank managers' risk perceptions in lending to micro, small, medium enterprises?

PurposeThe purpose of this research is to study the antecedents of risk perceptions of bank managers towards micro, small, medium enterprise (MSME) lending, in the situation of information asymmetry, where cognitive factors assume significance over organisational norms of lending.Design/methodology/approachThis study proposed and tested a conceptual model based on the factors identified from literature review and exploratory and quantitative study. Multinomial logistic regression technique is used for quantitative analysis.FindingsThe research postulates that information asymmetry, risk attitude, perceived trust and organizational norms have a significant relationship with branch managers' perceived risk in lending to MSMEs. The research emphasized that the risk attitude of managers and perceived trust moderate the relationship between information asymmetry and perceived risk. The findings and discussions enrich the knowledge about the alleviators of constraints to MSME funding in developing nations despite information asymmetry.Originality/valueAuthors have given holistic view on the risk perception in the financial decision-making process of bank lending. The research highlights the importance of cognitive factors in decreasing the negative impact of information asymmetry on risk perception.

Analysis of the Effect of Private Healthcare Financing on Poverty in Nigeria: Evidence from Edo State

Poverty is one of the problems that challenge economies in Africa. Though it is a complex phenomenon which requires efforts by different experts to reduce or eliminate, conventional wisdom posits that “health is wealth”. Health status is a component of human capital development which plays a fundamental role in the poverty and well-being of individuals and national economies. Paradoxically the cost of accessing quality healthcare is an important contributor to income poverty among low income households. Thus adequate healthcare financing mechanisms (public and private) are required to attain quality health outcomes. This study therefore investigates the adequacy or otherwise of the current means of private health care financing in Edo state of Nigeria and it employed the survey method and multinomial logistic regression technique. Results revealed that the dominant means of private health care financing in Edo state is “out of pocket” payments which has negative effect on the income of households. It therefore recommends the introduction of a more effective collective healthcare financing mechanism to mitigate the financial burden associated with out-of-pocket spending. Also funding should be provided for research and development of locally manufactured drugs with high local content to enhance the availability and affordability of effective drugs.

Factors that Influence the Individual Research Output of University Professors: The Case of Ecuador, Peru, and Colombia

The study analyzed the factors that influence the individual research output of university professors in Ecuador, Peru, and Colombia, using multinomial logistic regression technique. Results showed that previous research publications are the main factor that explains subsequent research production. Age, academic rank, time invested on research, resource allocation, recognition, and research leaders have also a direct effect on research output. A comprehensive model is proposed and implications for universities authorities are discussed.

Modeling Spatial Variation of Money Laundering Crime in Indonesia Using Geographically Weighted Multinomial Logistic Regression

This document presents the mapping of crime in Indonesia. Crime is a real threat that needs to be optimally handled which increases significantly and it is very detrimental to the country’s economy. The crime that occurred created another more serious crime, namely Money Laundering Crime. The impact of Money Laundering Crime is the flow of illicit fund through the payment system to financial service providers as well as providers of goods and services. The perpetrator of Money Laundering crime is a player who is shrewd and familiar with the state financial system, so that the flow of funds does not involve a region but will expand to other areas in a country. This study analyses the spatial variations in the factors causing the Money Laundering crime in Indonesia by using a Geographically Weighted Multinomial Logistic Regression (GWMLR) technique. In this document, the GWMLR method is used to model the Money Laundering Crime and its explanatory variables by considering the effect of geographic location using the Adaptive Kernel weighting function. The response variables were divided into four categories with nominal scales by combining the general criminal case and money laundering crime.

Economic Potential of Oil Palm Plantation Using Remote Sensing-Based Technology in Indonesia

Abstract. The productivity of Indonesia's palm oil was considered low when referring to the 14.6 million ha land area in 2019, with the production of national palm oil only reaching 3.2 tons of CPO/ha/year. The uses of remote sensing technology as a means of monitoring and supervising, were expected to increase oil palm production in line with productivity. The purpose of this study was to determine the economic potential based on oil palm plantation productivity, with and without using remote sensing-based technology, as well as other variables likely to affect productivity. Primary and secondary data collection methods were also used in this research. There were three quantitative methods being used in this study, namely (i) Multiple regression model with panel data, (ii) Data Envelopment Analysis (DEA) tool, and (iii) Multinomial logistic regression technique. The results showed that the generated economic potential from the utilization of the remote sensing model, had efficient opportunity value of 10.48, which was higher than the non-usage of the technology. Therefore, the main variables that affected productivity in this study, were fertilizer and labour. Keywords: Efficiency, oil-palm, remote sensing (spot 6), policy, binomial logistic

Crime distortion within the NYPD: a potential method for estimating crime misclassification within CompStat statistics

ABSTRACT Since the advent of CompStat in 1994, the NYPD contends that felony crimes have decreased throughout NYC, specifically within the seven-major felony index crime categories. Previous work from Eterno and Silverman has suggested that the scope of the decrease was duplicitous and exaggerated due to the department's widespread practice of crime distortion. However, previous research has never attempted to quantify this phenomenon. This study addresses a critical gap within the existing CompStat literature by attempting to capture the magnitude of crime distortion. Using secondary crime data from the NYPD, the current study will examine burglaries and the suspected downward misclassifications, occurring within all NYC precincts, at the aggregate level, during the period between 2000 and 2013. The goals of this research are to identify and summarize any precinct-level patterns of potential crime distortion using semi-parametric group-based trajectory modeling and multinomial logistic regression techniques.

Effectiveness of Micro-Credit Programs in Socio-Economic Development: A Study on Selected NGOs in Bangladesh

Non-governmental Organizations (NGOs) evolved in Bangladesh with the commitment of bringing positive social change and attaining balanced socio-economic development through the disbursement of micro-credit among the marginal people in society. The main intention of the study is to assess the impact of micro-credit on the overall socio-economic development of micro-credit receivers. The Chi-square (χ2) and Multinomial Logistic Regression techniques have applied on primary data collected from 260 beneficiaries of micro-credit of ASA and BRAC NGOs in the Mymensingh district of Bangladesh. The χ2 analysis shows that the amount of micro-credit taken has a significant positive impact on respondents’ type of home, sanitation facilities, ownership of other household assets, health service received, yearly income, expenditure, and savings, and participation in family’s decision making, whereas an insignificant relationship is observed for occupational status, children education, ownership of land and livestock, sources of drinking water, economic solvency, and social status of the respondents.

How Accumulated Wealth Affects Driving Reduction and Cessation.

Older adults can expect to live between 6 and 10 years after they give up driving, but driving reduction and cessation (DRC) are not equally experienced by all groups. Individual characteristics such as poor health, impaired vision, older age, and female gender are known to affect DRC. Using cumulative disadvantage theory as a guide, this study assessed the role played by wealth in DRC among older adults. Data from the National Health and Aging Trends Study were analyzed using multinomial logistic regression techniques. This allowed for the effect of each predictor on the odds of engagement in a given driving status (full driving, driving reduction [DR], and driving cessation [DC]) to be compared to each of the others. The final sample included 6,387 participants. After controlling for the effect of covariates, less wealth was associated with higher odds of DR compared to full driving, DC compared to full driving, and DC compared to DR. Confirming previous research, several other factors were also significantly related to driving status including age, health, vision, gender, race, education, relationship status, household size, and work status. The influence of wealth on driving status among older adults represents another disadvantage unequally distributed to some in older adulthood. Those with less wealth will have fewer resources to meet their mobility needs using alternatives and may already be facing additional financial constraints due to worse health and other challenges associated with lower socioeconomic status.

Geochemical Discrimination of Monazite Source Rock Based on Machine Learning Techniques and Multinomial Logistic Regression Analysis

Detrital monazite geochronology has been used in provenance studies. However, there are complexities in the interpretation of age spectra due to their wide occurrence in both igneous and metamorphic rocks. We use the multinomial logistic regression (MLR) and cross-validation (CV) techniques to establish a geochemical discrimination of monazite source rocks. The elemental abundance-based geochemical discrimination was tested by selecting 16 elements from granitic and metamorphic rocks. The MLR technique revealed that light rare earth elements (REEs), Eu, and some heavy REEs are important discriminators that reflect elemental fractionation during magmatism and/or metamorphism. The best model yielded a discrimination rate of ~97%, and the CV method validated this approach. We applied the discrimination model to detrital monazites from African rivers. The detrital monazites were mostly classified as granitic and of garnet-bearing metamorphic origins; however, their proportion of metamorphic origin was smaller than the proportion that was obtained by using the elemental-ratio-based discrimination proposed by Itano et al. in Chemical Geology (2018). Considering the occurrence of metamorphic rocks in the hinterlands and the different age spectra between monazite and zircon in the same rivers, a ratio-based discrimination would be more reliable. Nevertheless, our study demonstrates the advantages of machine-learning-based approaches for the quantitative discrimination of monazite.

Recognition of Sleep/Wake States analyzing Heart Rate, Breathing and Movement Signals.

This document presents an algorithm for a non-obtrusive recognition of Sleep/Wake states using signals derived from ECG, respiration, and body movement captured while lying in a bed. As a core mathematical base of system data analytics, multinomial logistic regression techniques were chosen. Derived parameters of the three signals are used as the input for the proposed method. The overall achieved accuracy rate is 84% for Wake/Sleep stages, with Cohen's kappa value 0.46. The presented algorithm should support experts in analyzing sleep quality in more detail. The results confirm the potential of this method and disclose several ways for its improvement.

Energy sources for cooking and its determinants in rural areas of Tanzania

Using traditional sources of energy for cooking in rural areas of Tanzania is associated with problems such as lung diseases, soil erosion and air pollution. This paper identifies the energy sources used for cooking and its determinants in rural areas of Tanzania. Cross-sectional survey design through questionnaire was used to collect data from 384 households living in rural areas of Njombe and Iringa regions in Tanzania. Multistage cluster sampling technique was employed to sample districts, wards and villages, while rural heads of households were sampled randomly using the fishbowl method to avoid biasness. Descriptive analysis such as frequencies and percentages were used. Multinomial Logistic Regression (MLR) techniques was used to estimate the parameters of factors determining the choices of energy sources for cooking in rural areas of Tanzania. The findings show that firewood is the main source of energy for cooking, followed by charcoal, Liquefied Petroleum Gas (LPG) and electricity in rural areas of Tanzania. The study shows that education, household size, occupation, income, and age of respondents determine the choices of energy sources for cooking. It can be concluded that, apart from improving income, other intervention such as family planning, reforestation programmes and promotion of the use of modern cooking stoves should be done to ensure sustainable development in rural areas of Tanzania

The role of neighborhoods in household food insufficiency: Considering interactions between physical disorder, low social capital, violence, and perceptions of danger

RationaleFood insecurity is a significant public health concern, with implications for community and individual health and well-being. Although a growing body of literature points to the role of neighborhoods in household food insecurity, studies using nationally representative samples to explore interactions between neighborhood risks – including violence and danger – are lacking. ObjectiveThe present study examines whether interactions between physical disorder, low social capital, and violence/danger in the neighborhood have significant implications for the risk of household food insufficiency using a large, nationally representative sample of U.S. children and their families. MethodData are from the 2016 National Survey of Children's Health, a survey of a cross-sectional weighted probability sample of U.S. children from 0 to 17 years of age. Multinomial logistic regression techniques were used to analyze the data. ResultsNeighborhood risk factors interacted to predict household food insufficiency, with the confluence of low social capital and violence/danger yielding the strongest effects. ConclusionsOur findings suggest that food hardship should be addressed within the context of neighborhood revitalization. The risk of food insufficiency among children and families in especially high-risk ecological contexts might be ameliorated with the provision of informal and formal sources of nutrition assistance and support.