Ordinal regression

About: Ordinal regression is a research topic. Over the lifetime, 1879 publications have been published within this topic receiving 65431 citations.

Influence of Subcontract Offering on the Performance of Manufacturing Micro and Small Enterprises in Kenya

Abstract: This study set out to investigate the influence of intensity of participation in subcontract offering on the performance of manufacturing micro and small enterprises (MSEs) in Kenya. The study used an exploratory research design targeting a population of 2450 MSEs from Kamukunji ‘JuaKali’ Association, Nairobi Kenya. A random sample of 180 firms returned 175 (97.2%) valid responses. Survey data was collected with a semistructured questionnaire through face-to-face interviews. A pilot test on 20 firms helped to improve the instrument while the Principal Component Analysis (PCA) method extracted the factors with reliability cut-off value of 0.70. Factors loadings that were less than 0.40 were discarded. Descriptive statistics presented the responses in means and standard deviations. To sharpen inferences, ordinal regression analysis was performed using the Polytomous Universal Model (PLUM) of SPSS for Windows 19 location-scale model. Response frequencies of firm performance, ordered in 5-part Likert-type categories, were positively skewed, thus,the negative log‐log link function was used. Model fitting information provided log likelihood ratio tests for the null hypothesis that the independent variable was statistically equal to zero. The study found that the intensity of participation in subcontract offering influences firm performance, positively and significantly.

Meta Ordinal Weighting Net For Improving Lung Nodule Classification

Abstract: The progression of lung cancer implies the intrinsic ordinal relationship of lung nodules at different stages—from benign to unsure then to malignant. This problem can be solved by ordinal regression methods, which is between classification and regression due to its ordinal label. However, existing convolutional neural network-based ordinal regression methods only focus on modifying classification head based on a randomly sampled mini-batch of data, ignoring the ordinal relationship resided in the data itself. In this paper, we propose a Meta Ordinal Weighting Network (MOW-Net) to explicitly align each training sample with a meta ordinal set (MOS) containing a few samples from all classes. During the training process, the MOW-Net learns a mapping from samples in MOS to corresponding class-specific weight. We further propose a meta cross-entropy loss to optimize the network in a meta-learning scheme. Experimental results demonstrate that the MOW-Net achieves better accuracy than the state-of-the-art ordinal regression methods, especially for the unsure class.

Regression modeling of ordinal data with nonzero baselines.

Abstract: Summary. This paper develops regression models for ordinal data with nonzero control response probabilities. The models are especially useful in dose-response studies where the spontaneous or natural response rate is nonnegligible and the dosage is logarithmic. These models generalize Abbott's formula, which has been commonly used to model binary data with nonzero background observations. We describe a biologically plausible latent structure and develop an EM algorithm for fitting the models. The EM algorithm can be implemented using standard software for ordinal regression. A toxicology data set where the proposed model fits the data but a more conventional model fails is used to illustrate the methodology.

Measuring risk with ordinal variables

Abstract: In this paper we propose a novel approach to measure risks, when the data available are expressed in an ordinal scale. As a result we obtain a new index of risk bounded between 0 and 1, that leads to a risk ordering that is consistent with a stochastic dominance approach. The proposed measure, being non parametric, can be applied to a wide range of problems, where data are ordinal and where a point estimate of risk is needed. We also provide a method to calculate confidence intervals for the proposed risk measure, in a Bayesian non parametric framework. In order to evaluate the actual performance of what we propose, we analyse a database provided by a telecommunication company, with the final aim of measuring operational risks, starting from a self-assessment questionnaire.