• Communications for Statistical Applications and Methods
• /
• v.12 no.2
• /
• pp.313-322
• /
• 2005

In this paper, we discuss suppression for logistic regression model. Suppression for linear regression model was defined as the relationship among sums of squared for regression as well as correlation coefficients of. variables. Since it is not common to obtain simple correlation coefficient for binary response variable of logistic model, we consider cumulative logistic models with multinomial and ordinal response variables rather than usual logistic model. As number of category of a response variable for the cumulative logistic model gets collapsed into binary, it is found that suppressions for these logistic models are changed. These suppression results for cumulative logistic models are discussed and compared with those of linear model.

• Journal of Environmental Science International
• /
• v.25 no.4
• /
• pp.603-613
• /
• 2016

Fog may have a significant impact on road conditions. In an attempt to improve fog predictability in Jeju, we conducted machine learning with various data mining techniques such as tree models, conditional inference tree, random forest, multinomial logistic regression, neural network and support vector machine. To validate machine learning models, the results from the simulation was compared with the fog data observed over Jeju(184 ASOS site) and Gosan(185 ASOS site). Predictive rates proposed by six data mining methods are all above 92% at two regions. Additionally, we validated the performance of machine learning models with WRF (weather research and forecasting) model meteorological outputs. We found that it is still not good enough for operational fog forecast. According to the model assesment by metrics from confusion matrix, it can be seen that the fog prediction using neural network is the most effective method.

• Korea Trade Review
• /
• v.47 no.2
• /
• pp.69-88
• /
• 2022

This paper aims to predict Busan's regional product and employment using the logistic regression models and machine learning models. The following are the main findings of the empirical analysis. First, the OLS regression model shows that the main industries such as electricity and electronics, machine and transport, and finance and insurance affect the Busan's income positively. Second, the binomial logistic regression models show that the Busan's strategic industries such as the future transport machinery, life-care, and smart marine industries contribute on the Busan's income in large order. Third, the multinomial logistic regression models show that the Korea's main industries such as the precise machinery, transport equipment, and machinery influence the Busan's economy positively. And Korea's exports and the depreciation can affect Busan's economy more positively at the higher employment level. Fourth, the voting ensemble model show the higher predictive power than artificial neural network model and support vector machine models. Furthermore, the gradient boosting model and the random forest show the higher predictive power than the voting model in large order.

• International Journal of Computer Science & Network Security
• /
• v.21 no.1
• /
• pp.1-5
• /
• 2021

Cyberbullying is a problem that is faced in many cultures. Due to their popularity and interactive nature, social media platforms have also been affected by cyberbullying. Social media users from Arab countries have also reported being a target of cyberbullying. Machine learning techniques have been a prominent approach used by scientists to detect and battle this phenomenon. In this paper, we compare different machine learning algorithms for their performance in cyberbullying detection based on a labeled dataset of Arabic YouTube comments. Three machine learning models are considered, namely: Multinomial Naïve Bayes (MNB), Complement Naïve Bayes (CNB), and Linear Regression (LR). In addition, we experiment with two feature extraction methods, namely: Count Vectorizer and Tfidf Vectorizer. Our results show that, using count vectroizer feature extraction, the Logistic Regression model can outperform both Multinomial and Complement Naïve Bayes models. However, when using Tfidf vectorizer feature extraction, Complement Naive Bayes model can outperform the other two models.

• Earthquakes and Structures
• /
• v.26 no.2
• /
• pp.87-96
• /
• 2024

This paper proposes a logistic multinomial regression approach to model the spatial cross-correlation of damage probabilities among different damage states in an expanded transportation network. Utilizing Bayesian theory and the multinomial logistic model, we analyze the damage states and probabilities of bridges while incorporating damage correlation. This correlation is considered both between bridges in a network and within each bridge's damage states. The correlation model of damage probabilities is applied to the seismic assessment of a portion of Tehran's transportation network, encompassing 26 bridges. Additionally, we introduce extra daily traffic time (EDTT) as an operational parameter of the transportation network and employ the shortest path algorithm to determine the path between two nodes. Our results demonstrate that incorporating the correlation of damage probabilities reduces the travel time of the selected network. The average decrease in travel time for the correlated case compared to the uncorrelated case, using two selected EDTT models, is 53% and 71%, respectively.

• The Journal of Asian Finance, Economics and Business
• /
• v.8 no.12
• /
• pp.391-397
• /
• 2021

One of the sustainable development goals is to promote good health and well-being for all people. Child health is a top priority since their health issues can have a detrimental impact on human capital development, which is a critical input for the growth model. This paper applies the health production model to explore the determinants that influence the health of children under the age of five. The results of a survey of 203 households in Ho Chi Minh City, Vietnam, were examined. Child health is measured using anthropometric indicators such as weight-for-age, height-for-age, and weight-for-height (ZWFH). Three separate multinomial logistic models are regressed to examine the drivers of child health as proxied by z-score weight for age, z-score height for age, and z-score weight for height. The significance of input variables relating to a child's attributes, household, and environment was validated by the findings. The inclusion of overweight besides under-nourished indexes is novel because it reflects the current trend of child over-nutrition. The findings of the study highlight the importance of a wide range of initiatives to enhance child health. Moreover, the genetic effect is found to be crowded out by environmental and household factors. The finding verifies that despite their parents' moderate height, the future generation of Vietnamese can achieve the desired height.

• The Journal of Engineering Geology
• /
• v.31 no.1
• /
• pp.31-42
• /
• 2021

Database of Cut-slope management system (CSMS) has been constructed based on investigations of all slopes on the roads of the whole country. The investigation data is documented by human, so it is inevitable to avoid human-error such as missing-data and incorrect entering data into computer. The goal of this paper is constructing a prediction model based on several machine-learning algorithms to solve those imperfection problems of the CSMS data. First of all, the character-type data in CSMS data must be transformed to numeric data. After then, two algorithms, i.g., multinomial logistic regression and deep-neural-network (DNN), are performed, and those prediction models from two algorithms are compared. Finally, it is identified that the accuracy of DNN-model is better than logistic model, and the DNN-model will be utilized to improve data-quality.

• Journal of the Korean Data and Information Science Society
• /
• v.28 no.3
• /
• pp.685-695
• /
• 2017

The traditional mixture of experts (ME) modeled the gate network using a certain parametric function. However, if the assumed parametric function does not properly reflect the true nature, the prediction strength of ME would become weak. For example, the parametric ME often uses logistic or multinomial logistic models for the network model. However, this could be very misleading if the true nature of the data is quite different from those models. Although, in this case, we may develop more flexible parametric models by extending the model at hand, we will never be free from such misspecification problems. In order to alleviate such weakness of the parametric ME, we propose to use the semi-parametric mixture of experts (SME) in which the gate network is estimated in a non-parametrical way. Based on this, we compared the performance of the SME with those of ME and neural networks via several simulation experiments and real data examples.

• Asian Pacific Journal of Cancer Prevention
• /
• v.16 no.4
• /
• pp.1627-1631
• /
• 2015

Background: Stage at diagnosis is one of the most important prognostic factors of breast cancer survival. Because in the breast cancer case this may vary with socioeconomic characteristics, this study was performed to recognize the relationship between demographic and socioeconomic factors with stage at diagnosis in Iran. Materials and Methods: This cross-sectional, descriptive study conducted on 526 patients suffering from breast cancer and registered in Cancer Research Center of Shahid Beheshti University of Medical Sciences from 2008 to 2013. A reliable and valid questionnaire about family levels of socioeconomic status filled in by interviewing the patients via phone. For analyzing the data, Multinomial logistic regression, Kendal tau-b correlation coefficient and Contingency Coefficient tests were executed by SPSS22. Economic status, educational attainment of patient and household head and/or a combination of these were considered as parameters for socioeconomic status. First, the relationship between stage at diagnosis and demographic and socioeconomic status was assessed in univariate analysis then these relationships assessed in two different models of multinomial logistic regression. Results: The mean age of the patients was 48.3 (SD=11.4). According to the results of this study, there were significant relationships between stage at diagnosis of breast cancer with patient education (p=0.011), living place (p=0.044) and combined socioeconomic status (p=0.024). These relationships persisted in multiple multinomial logistic regressions. Other variables, however, had no significant correlation. Conclusions: Patient education, combined socioeconomic status and living place are important variables in stage at diagnosis of breast cancer in Iranian women. Interventions have to be applied with the aim of raising women's accessibility to diagnostic and medical facilities and also awareness in order to reducing delay in referring. In addition, covering breast cancer screening services by insurance is recommended.

• Culinary science and hospitality research
• /
• v.12 no.3 s.30
• /
• pp.49-60
• /
• 2006

The purpose of this study was to measure the factors influencing on the perception of helpfulness of marking the country of origin in predicting the quality and safety of pork. A total of 239 questionnaires were completed. A multinomial logit model is specified in order to estimate which factors influence the probability that a consumer perceives the country of origin as helpful in assessing food quality and food safety. The estimations were carried out using the logistic procedure of SAS. The results are as follows. The proportional odds assumptions of models were not violated at p<0.05. The effects of age, income, children, occupation and respondents informed on the importance of the country of origin in pork quality model were statistically significant. The effects of age, children, occupation and trust on the importance of the country of origin in pork safety model were statistically significant. The results from this study could be useful in developing marketing and health promotion strategies as well as government trade policies.