Data classification with binary response through the Boosting algorithm and logistic regression

Menezes, Fortunato Silva de; Liska, Gilberto Rodrigues; Cirillo, Marcelo Ângelo; Vivanco, Mario Javier Ferrua

doi:10.1016/j.eswa.2016.08.014

Cited by 63 publications

(32 citation statements)

References 16 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…In order to discriminate patients with presence/absence of coronary heart disease, Menezes et al (2016) has adjusted the Boosting algorithm and has reached the best rates of accuracy, sensitivity, specificity, false positive rates and false negative rates as compared to the rates obtained by the logistic regression model with its estimated parameters via maximum likelihood.…”

Section: Resultsmentioning

confidence: 99%

“…However, it is widely known that a combination of opinions leads to a more coherent decision than one taken individually. Due to this fact, computer researchers have proposed methods of classification based on the combination of classifiers with the intent to reproduce the ability of humans on machines to improve network performance on some tasks, which makes the classification stand as the most important ones (MENEZES,2016) . (Taki,2018) states that Machine Learning Techniques have been successfully applied in a wide range in agriculture.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Machine Learning techniques in muliclass problems with application in sensorial analysis

Oliveira

Menezes

Cirillo

et al. 2019

Concurrency and Computation

Self Cite

View full text Add to dashboard Cite

Summary Automatic classification methods have been developed in the area of Machine Learning to facilitate the categorization of data. Among the most successful methods are Boosting and Bagging. While Bagging works by combining fit classifiers into the bootstrap samples, Boosting works by sequentially applying a sorting algorithm to reweigh versions of the training dataset, giving more weight to the erroneously classified observations in the previous step. These classifiers are characterized by satisfactory results, low computational cost, and simplicity of implementation. Given these characteristics, there is an interest in verifying the performance of these automatic methods compared to the classical methods of classification in Statistics such as Linear and Quadratic Discriminant Analysis. To compare these techniques, we have used the classification error rates of the models to improve the confidence in the use of Boosting and Bagging methods in more complex classification problem. This study applies these techniques to real and simulated data that have been composed of more than two categories in the response variable. This investigation stimulates the implementation of Boosting and Bagging, by assigning an application in Sensory Analysis. We have concluded that the automatic methods have an optimal classification performance, showing lower error rates compared to the Linear and Quadratic Discriminant Analysis in the tested applications.

show abstract

Section: Resultsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Machine Learning techniques in muliclass problems with application in sensorial analysis

Oliveira

Menezes

Cirillo

et al. 2019

Concurrency and Computation

Self Cite

View full text Add to dashboard Cite

show abstract

“…Logistic regression is a powerful tool used for investigating socio-economic factors (Smart, Harrison 2017) when the relationship between variables is not linear. One of the advantages of using the logistic regression model is to obtain a relationship between the probability of the oc- currence of a dependent and independent variable (de Menezes et al 2017). The model of multiple logistic regression is presented below (1) (Pandya et al 2014):…”

Section: Methodsmentioning

confidence: 99%

Factors Affecting Personal Customers’ Trust in Traditional Banking: Case of the Baltics

Skvarciany

Jurevičienė

2017

Journal of Business Economics and Management

View full text Add to dashboard Cite

Nobody disputes that trust is an important issue in choosing a financial service provider, especially in the area of new forms of banking. The goal of this paper is to assess the most important determinants of trust in traditional banking. The study was conducted in the Baltics and personal customers had to rank the distinguished factors. Using correlation analysis and binary logistic regression model it was found that the most significant factor influencing trust in all countries is provided information by the bank. In addition, in Lithuania -bank's characteristics, in Latvia -customers' risk perception and bank's characteristics, in Estonia -respondents' experience of cooperation with a bank were highlighted as significant. The following measures of fit are used in order to describe the created logistic models: contingency table test, Nagelkerke pseudo-R 2 , Pearson chisquare test, Wald test. However, there is a limitation -the survey was conducted online. Nevertheless, as Internet penetration rate is high enough in investigated countries (from 76 percent in Latvia to 91 percent in Estonia), survey results can be adapted for at least seventy-five percent of each country's population. The findings have implications on the development of the strategy and the policy of commercial banks.

show abstract

“…The value of variable P can be between 0 and +∞, hence the value of Ln (P) is between -∞ and +∞. 34 At first, the coefficients are estimated by the statistical techniques such as Maximum Likelihood and then the LR function is obtained. In LR modeling, there are different methods for selection and entering variables to the model.…”

Section: Model Outputmentioning

confidence: 99%

A Predictive Model for Assessment of Successful Outcome in Posterior Spinal Fusion Surgery

Babaee

Soleimani

Zali

et al. 2017

Int Clin Neurosci J

View full text Add to dashboard Cite

Background: Low back pain is a common problem in many people. Neurosurgeons recommend posterior spinal fusion (PSF) surgery as one of the therapeutic strategies to the patients with low back pain. Due to the high risk of this type of surgery and the critical importance of making the right decision, accurate prediction of the surgical outcome is one of the main concerns for the neurosurgeons. Methods: In this study, 12 types of multi-layer perceptron (MLP) networks and 66 radial basis function (RBF) networks as the types of artificial neural network methods and a logistic regression (LR) model created and compared to predict the satisfaction with PSF surgery as one of the most well-known spinal surgeries. Results:The most important clinical and radiologic features as twenty-seven factors for 480 patients (150 males, 330 females; mean age 52.32 ± 8.39 years) were considered as the model inputs that included: age, sex, type of disorder, duration of symptoms, job, walking distance without pain (WDP), walking distance without sensory (WDS) disorders, visual analog scale (VAS) scores, Japanese Orthopaedic Association (JOA) score, diabetes, smoking, knee pain (KP), pelvic pain (PP), osteoporosis, spinal deformity and etc. The indexes such as receiver operating characteristicarea under curve (ROC-AUC), positive predictive value, negative predictive value and accuracy calculated to determine the best model. Postsurgical satisfaction was 77.5% at 6 months follow-up. The patients divided into the training, testing, and validation data sets.

show abstract

Data classification with binary response through the Boosting algorithm and logistic regression

Cited by 63 publications

References 16 publications

Machine Learning techniques in muliclass problems with application in sensorial analysis

Machine Learning techniques in muliclass problems with application in sensorial analysis

Factors Affecting Personal Customers’ Trust in Traditional Banking: Case of the Baltics

A Predictive Model for Assessment of Successful Outcome in Posterior Spinal Fusion Surgery

Contact Info

Product

Resources

About