A corporate credit rating model using multi-class support vector machines with an ordinal pairwise partitioning approach

Predicting corporate credit-rating using statistical and artificial intelligence (AI) techniques has received considerable research attention in the literature. In recent years, multi-class support vector machines (MSVMs) have become a very appealing machine-learning approach due to their good performance. Until now, researchers have proposed a variety of techniques for adapting support vector machines (SVMs) to multi-class classification, since SVMs were originally devised for binary classification. However, most of them have only focused on classifying samples into nominal categories; thus, the unique characteristic of credit-rating - ordinality - seldom has been considered in the proposed approaches. This study proposes a new type of MSVM classifier (named OMSVM) that is designed to extend the binary SVMs by applying an ordinal pairwise partitioning (OPP) strategy. Our model can efficiently and effectively handle multiple ordinal classes. To validate OMSVM, we applied it to a real-world case of bond rating. We compared the results of our model with those of conventional MSVM approaches and other AI techniques including MDA, MLOGIT, CBR, and ANNs. The results showed that our proposed model improves the performance of classification in comparison to other typical multi-class classification techniques and uses fewer computational resources.     

Data envelopment analysis for evaluating the efficiency of genetic algorithms on solving the vehicle routing problem with soft time windows

This study proposes an alternative to the conventional empirical analysis approach for evaluating the relative efficiency of distinct combinations of algorithmic operators and/or parameter values of genetic algorithms (GAs) on solving the pickup and delivery vehicle routing problem with soft time windows (PDVRPSTW). Our approach considers each combination as a decision-making unit (DMU) and adopts data envelopment analysis (DEA) to determine the relative and cross efficiencies of each combination of GA operators and parameter values on solving the PDVRPSTW. To demonstrate the applicability and advantage of this approach, we implemented a number of combinations of GA’s three main algorithmic operators, namely selection, crossover and mutation, and employed DEA to evaluate and rank the relative efficiencies of these combinations. The numerical results show that DEA is well suited for determining the efficient combinations of GA operators. Among the combinations under consideration, the combinations using tournament selection and simple crossover are generally more efficient. The proposed approach can be adopted to evaluate the relative efficiency of other meta-heuristics, so it also contributes to the algorithm development and evaluation for solving combinatorial optimization problems from the operational research perspective.     

Direction-based adaptive data propagation for heterogeneous sensor mobility

We consider sensor networks where the sensor nodes are attached on entities that move in a highly dynamic, heterogeneous manner. To capture this mobility diversity we introduce a new network parameter, the direction-aware mobility level, which measures how fast and close each mobile node is expected to get to the data destination (the sink). We then provide local, distributed data dissemination protocols that adaptively exploit the node mobility to improve performance. In particular, “high” mobility is used as a low cost replacement for data dissemination (due to the ferrying of data), while in the case of “low” mobility either (a) data propagation redundancy is increased (when highly mobile neighbors exist) or (b) long-distance data transmissions are used (when the entire neighborhood is of low mobility) to accelerate data dissemination toward the sink. An extensive performance comparison to relevant methods from the state of the art demonstrates significant improvements, i.e. latency is reduced by even four times while keeping energy dissipation and delivery success at very satisfactory levels.     

Evaluation of a workplace technology for mental health assessment: A meaning-making process

This paper describes the results of a summative evaluation conducted to explore the usability and acceptability of an automated telephone-based technology, designed and developed as a screening tool for mental health disorders in the workplace. The system screens employees for symptoms and subsequently provides educational information and referrals to improve mental health and enhance productivity. The qualitative evaluation was a sub-study of a larger randomized controlled trial (RCT). We conducted in-depth interviews with 15 individuals who participated in the RCT and had used the system. A phenomenological theoretical framework was utilized to analyze the narrative data. Results indicated that participants had mostly positive experience with the system and perceived it as a confidential program that allowed them to reflect on their symptoms without inhibition, fear of judgment or embarrassment. We also learned that asking mere questions about mental health symptoms can be a positive, instructive and possibly a restorative experience. There is a likelihood that for many participants, exposure to questions about their emotional and mental distress validated their feelings about their symptoms and was sufficiently compelling to promote help-seeking behavior. This evaluation study demonstrates that computers can be utilized in the workplace as effective screening and educational tools to improve employees’ mental health.     

Inter-rater reliability of output measures for a posture matching assessment approach: a pilot study with food service workers

Despite the ongoing health problem of repetitive strain injuries, there are few tools currently available for ergonomic applications evaluating cumulative loading that have well-documented evidence of reliability and validity. The purpose of this study was to determine the inter-rater reliability of a posture matching based analysis tool (3DMatch, University of Waterloo) for predicting cumulative and peak spinal loads. A total of 30 food service workers were each videotaped for a 1-h period while performing typical work activities and a single work task was randomly selected from each for analysis by two raters. Inter-rater reliability was determined using intraclass correlation coefficients (ICC) model 2,1 and standard errors of measurement for cumulative and peak spinal and shoulder loading variables across all subjects. Overall, 85.5% of variables had moderate to excellent inter-rater reliability, with ICCs ranging from 0.30–0.99 for all cumulative and peak loading variables. 3DMatch was found to be a reliable ergonomic tool when more than one rater is involved.     

Function simulation,graph grammars and colourings

We prove that to any partial function ? defined on a finite set, there corresponds an infinite class of graphs that could be generated by a graph grammar such that each graph in the class represents the function in the sense that evaluation of the function at any point x of its domain can be simulated by finding the unique extension of a partial vertex colouring of the graph specified by x. We show that in the proposed setup, generating such simulator graphs as well as finding the colouring extensions can be computed effectively in polynomial time. We also discuss some applications of this scenario in producing instances of the graph colouring problem near its phase transition that can be applied in a cryptographic setting.     

Preventing user interface disasters

Abstract

This article defines a quantitative goal that is cheap to measure for the usability of a business application system for casual users. The article also describes a cost-effective method for attaining the goal. The goal is to eliminate all user interface disasters (UIDs) in a given system. UIDs are usability problems that seriously annoy users, or prevent them from accomplishing their work without help from a human being. The method consists of series of simple user tests without audio or video recording, and with little analysis after each user test. The article concludes by describing Baltica's results of applying the method to a medium-size business application for casual users.     

面向星球探测的岩石三维表面评估技术

针对星球探测任务中的岩石采样问题,利用双目立体视觉装置采集不同岩石图像,采用立体视觉算法获得岩石表面的三维点云数据。以三角剖分后生成的三角片法向量夹角为基本元素,提出基于C均值聚类的岩石表面评估算法,对岩石平整三维表面进行评估计算,并用不同颜色加以区分。仿真结果证明了该算法的有效性。     

一种用于隐写测评的图像纹理复杂度估计方法

在评估图像隐写算法安全性及隐写分析性能时,隐写嵌入率等参数是重要的评价指标,而图像纹理差异对评价的影响很容易被忽略。为此,基于图像高维统计模型,提出一种新的针对隐写应用的图像纹理复杂度度量方法,用于衡量图像间的纹理差异。根据纹理复杂度对图像进行分类,单独在每一类不同纹理复杂度的图像集上评估隐写分析的检测性能。实验结果表明,该方法能有效地评价隐写的安 全性。     

基于J48决策树算法的水质评价方法

提出一种以J48决策树算法为基础的水质评价方法。通过分析K最近邻法、贝叶斯分类法、向量空间模型法等常用的水质评价方法,针对某地区2008年的水质监测数据,采用J48决策树算法和单因子污染指数法对该地区水质进行计算,并对水质分类数据进行分类、属性清理和J48决策树参数的选择,以此建立模型。实验结果表明,该方法的交叉验证率约为95%,具有较好的分类评估效果。