Improving experimental studies about ensembles of classifiers for bankruptcy prediction and credit scoring

Previous studies about ensembles of classifiers for bankruptcy prediction and credit scoring have been presented. In these studies, different ensemble schemes for complex classifiers were applied, and the best results were obtained using the Random Subspace method. The Bagging scheme was one of the ensemble methods used in the comparison. However, it was not correctly used. It is very important to use this ensemble scheme on weak and unstable classifiers for producing diversity in the combination. In order to improve the comparison, Bagging scheme on several decision trees models is applied to bankruptcy prediction and credit scoring. Decision trees encourage diversity for the combination of classifiers. Finally, an experimental study shows that Bagging scheme on decision trees present the best results for bankruptcy prediction and credit scoring.     

The optimal strategy of a tree classifier

This paper deals with the decision rules of a tree classifier for performing the classification at each nonterminal node, under the assumption of complete probabilistic information. For given tree structure and feature subsets to be used, the optimal decision rules (strategy) are derived which minimize the overall probability of misclassification. The primary result is illustrated by an example.     

Subsampling techniques and the Jackknife methodology in the estimation of the extremal index

For a sequence of independent, identically distributed random variables any limiting point process for the time normalized exceedances of high levels is a Poisson process. However, for stationary dependent sequences, under general local and asymptotic dependence restrictions, any limiting point process for the time normalized exceedances of high levels is a compound Poisson process, i.e., there is a clustering of high exceedances, where the underlying Poisson points represent cluster positions, and the multiplicities correspond to the cluster sizes. For such classes of stationary sequences there exists the extremal indexθ, 0?θ?1, directly related to the clustering of exceedances of high values. The extremal index θ is equal to one for independent, identically distributed sequences, i.e., high exceedances appear individually, and θ>0 for “almost all” cases of interest. The estimation of the extremal index through the use of the Generalized Jackknife methodology, possibly together with the use of subsampling techniques, is performed. Case studies in the fields of environment and finance will illustrate the performance of the new extremal index estimator comparatively to the classical one.     

混合浮动车交通状态估计的样本数量研究

在保证GPS浮动车获取的交通信息准确可靠的前提下,如何确定浮动车数量是浮动车信息采集技术应用中的一个重要问题。在考虑数据有效性、浮动车运行速度和采样时间间隔等因素的基础上,重点考虑了浮动车样本的混合情况,分析了混合比例系数对样本数量的影响,通过对不同种类浮动车的计算模型进行融合,建立了面向道路网的浮动车样本数量计算模型。仿真实验表明：与前人模型相比,该模型更具合理性,并极大地提高了模型的实际应用能力。     

Determination of sample size in a rare event simulation method

A technique is proposed for evaluating the probability of aircraft collision by the importance sampling technique. Upper limits for the size of an auxiliary sample are found, which ensure the prescribed relative accuracy of probability evaluation. Theoretical conclusions are confirmed by numerical experiments. __________ Translated from Kibernetika i Sistemnyi Analiz, No. 1, pp. 76–86, January–February 2006.     

Evaluation of incremental learning algorithms for HMM in the recognition of alphanumeric characters

We present an evaluation of incremental learning algorithms for the estimation of hidden Markov model (HMM) parameters. The main goal is to investigate incremental learning algorithms that can provide as good performances as traditional batch learning techniques, but incorporating the advantages of incremental learning for designing complex pattern recognition systems. Experiments on handwritten characters have shown that a proposed variant of the ensemble training algorithm, employing ensembles of HMMs, can lead to very promising performances. Furthermore, the use of a validation dataset demonstrated that it is possible to reach better performances than the ones presented by batch learning.     

飞机航段油耗区间估计最小样本量分析

为解决基于QAR数据进行固定机型固定航段油耗区间估计时如何确定QAR数据样本量的问题,提出一种飞机航段油耗区间估计最小样本量的求解及检验方法。根据固定机型固定航段油耗样本满足N(μ,σ2)正态分布,在满足给定估计精度和显著水平的条件下,推导航段油耗区间估计应满足的最小样本量,利用χ2拟合检验法对最小样本量进行验证,得到最小样本量成立的最小概率要求。通过实例验证了该方法的有效性,为航空公司进行固定机型固定航段油耗估算提供了可行的参考方法。     

Enhancing the effectiveness of Ant Colony Decision Tree algorithms by co-learning

Data mining and visualization techniques for high-dimensional data provide helpful information to substantially augment decision-making. Optimization techniques provide a way to efficiently search for these solutions. ACO applied to data mining tasks – a decision tree construction – is one of these methods and the focus of this paper. The Ant Colony Decision Tree (ACDT) approach generates solutions efficiently and effectively but scales poorly to large problems. This article merges the methods that have been developed for better construction of decision trees by ants. The ACDT approach is tested in the context of the bi-criteria evaluation function by focusing on two problems: the size of the decision trees and the accuracy of classification obtained during ACDT performance. This approach is tested in co-learning mechanism, it means agents–ants can interact during the construction decision trees via pheromone values. This cooperation is a chance of getting better results. The proposed methodology of analysis of ACDT is tested in a number of well-known benchmark data sets from the UCI Machine Learning Repository. The empirical results clearly show that the ACDT algorithm creates good solutions which are located in the Pareto front. The software that implements the ACDT algorithm used to generate the results of this study can be downloaded freely from http://www.acdtalgorithm.com.     

Improving the state estimation for optimal control of stochastic processes subject to multiplicative noise

Computational models for the neural control of movement must take into account the properties of sensorimotor systems, including the signal-dependent intensity of the noise and the transmission delay affecting the signal conduction. For this purpose, this paper presents an algorithm for model-based control and estimation of a class of linear stochastic systems subject to multiplicative noise affecting the control and feedback signals. The state estimator based on Kalman filtering is allowed to take into account the current feedback to compute the current state estimate. The optimal feedback control process is adapted accordingly. The resulting estimation error is smaller than the estimation error obtained when the current state must be predicted based on the last feedback signal, which reduces variability of the simulated trajectories. In particular, the performance of the present algorithm is good in a range of feedback delay that is compatible with the delay induced by the neural transmission of the sensory inflow.     

仿射非线性飞机运动方程的RHO优化控制

飞机运动为非线性形式且扰动有可能极大,在变化较大飞行范围内使用常规飞控系统控制这种对象会产生误差.阐述一种逐点线性化后退区间优化控制(RHO)算法,有效解决小迎角范围内飞机仿射非线性运动控制问题.首先根据飞机仿射非线性运动方程特点,提出采用逐点线性化方法,在采样点上将仿射非线性方程变为时变线性方程,以此为控制对象提出RHO控制概念,并进一步推导出两个微分Riccati(DRE)方程,分析了DRE的解法,从而计算出在线控制指令.最后结合国内某型号飞机气动数据对此算法进行了仿真,证实算法的可行性和实时性.     

Index based optimal anatomy of a metamorphic manipulator for a given task

This paper introduces an approach for the determination of the best anatomy of a metamorphic manipulator for a given task at a given location. The location of the task is determined by maximizing the performance of a current industrial fixed anatomy robot. Two types of tasks are considered: a point to point task and a path following task, where in the first case the approximated minimum of the manipulability index is formed along the task points and in the second case the approximated minimum of the manipulator velocity ratio is formed along the line segments. These indexes are maximized in order to determine the best anatomy for the task. The proposed approach is tested and the results show that the determined best anatomy for each type of task acquired higher performance than the respective one achieved by the fixed anatomy manipulator.     

On the optimal control of a hydroelectric power plant

A class of optimal control problems with phase restrictions is investigated whose performance index and state equation are linear in the control variable. First some necessary conditions for optimality are proved and then they are used to get the optimal solution.     

模幂与点乘m_ary算法中窗口大小的最优化估计*

提出一种新大数模幂与点乘m_ary算法中窗口大小的最优化估计方法。该方法不同于传统的暴力搜寻方法,也不同于在窗口的取值范围内通过逐一测试程序来获得最优窗口大小的方法。其基于以下理论分析：模幂 m_ary算法的基本运算为大数乘法,其中包括大数平方算法和一般大数乘法;椭圆曲线加密算法中点乘的m_ary算法步骤与模幂的m_ary算法相同,后者的基本运算为倍乘和加法。根据m_ary算法的基本运算的调用次数,推算出了最优窗口大小的估计公式。通过实验对m_ary算法进行实现,并测试分析了根据估计公式计算出窗口大小的算     

模幂与点乘m_ary算法中窗口大小的最优化估计

提出一种新大数模幂与点乘m_ary算法中窗口大小的最优化估计方法.该方法不同于传统的暴力搜寻方法,也不同于在窗口的取值范围内通过逐一测试程序来获得最优窗口大小的方法.其基于以下理论分析:模幂 m_ary算法的基本运算为大数乘法,其中包括大数平方算法和一般大数乘法;椭圆曲线加密算法中点乘的m_ary算法步骤与模幂的m_ary算法相同,后者的基本运算为倍乘和加法.根据m_ary算法的基本运算的调用次数,推算出了最优窗口大小的估计公式.通过实验对m_ary算法进行实现,并测试分析了根据估计公式计算出窗口大小的算法实现时间效率与理论分析基本吻合.     

Comparison of fusion methods from the abstract level and the rank level in a dispersed decision-making system

Issues related to decision making based on dispersed knowledge are discussed in the paper. A dispersed decision-making system, which was proposed by the authors in previous articles, is used in this paper. In the system, a process of combining classifiers into coalitions with a negotiation stage is realized. The novelty that is proposed in this article involves the use of six different methods of conflict analysis that are known from the literature.The main purpose of the tests, which were performed, was to compare the methods from the two groups – the abstract level and the rank level. An additional aim was to investigate the efficiency of the fusion methods used in a dispersed system with a dynamic structure with the efficiency that is obtained when no structure is used. Conclusions were drawn that, in most cases, the use of a dispersed system improves the efficiency of inference.     

Determination of optimal dimensionality in statistical pattern classification

The effect of learning sample size on the optimal pattern recognition dimensionality is considered. Some procedures for determination of the optimal dimensionality are described and compared by a simulation method.     

On the role of optimality functions in numerical optimal control

We present a survey of optimality conditions in optimality function form and discuss their role in establishing that discretized optimal control problems are consistent approximations to the original optimal control problems.     

Wavelet based approximations in the optimal control of parabolic problems

We consider optimal control of a bilinear parabolic equation. The determination of such control requires to minimize a given energy performance measure. The performance measure of the system is taken as a combination of its modified total energy and the penalty term describing the approach used in the control process. Using an appropriate transformation modal expansion, the optimal control of a distributed parameter system (DPS) is simplified into the optimal control of a bilinear time-varying lumped parameter system (LPS). A computational efficient formulation to evaluate the optimal trajectory and control of the system is determined. It is based on the parametrization of the state and control variables by using finite wavelets. Numerical examples are provided to demonstrate the applicability and the efficiency of the proposed method and the results are quite satisfactory.     

Estimation of distributions involving unobservable events: the case of optimal search with unknown Target Distributions

We consider the problem of estimating the parameters of a distribution when the underlying events are themselves unobservable. The aim of the exercise is to perform a task (for example, search a web-site or query a distributed database) based on a distribution involving the state of nature, except that we are not allowed to observe the various “states of nature” involved in this phenomenon. In particular, we concentrate on the task of searching for an object in a set of N locations (or bins) {C ₁, C ₂,…, C _N }, in which the probability of the object being in the location C _i is p _i , where P = [p ₁, p ₂,…, p _N ] ^T is called the Target Distribution. Also, the probability of locating the object in the bin within a specified time, given that it is in the bin, is given by a function called the Detection function, which, in its most common instantiation, is typically, specified by an exponential function. The intention is to allocate the available resources so as to maximize the probability of locating the object. The handicap, however, is that the time allowed is limited, and thus the fact that the object is not located in bin C _i within a specified time does not necessarily imply that the object is not in C _i . This problem has applications in searching large databases, distributed databases, and the world-wide web, where the location of the files sought for are unknown, and in developing various military and strategic policies. All of the research done in this area has assumed the knowledge of the {p _i }. In this paper we consider the problem of obtaining error bounds, estimating the Target Distribution, and allocating the search times when the {p _i } are unknown. To the best of our knowledge, these results are of a pioneering sort - they are the first available results in this area, and are particularly interesting because, as mentioned earlier, the events concerning the Target Distribution, in themselves, are unobservable.

基于过度自信的资本市场多期最优激励契约研究

在考虑经纪人过度自信变化的基础上,研究投资者与经纪人之间的多期激励契约．从行为金融学角度,建立经纪人过度自信的动态模型并指出其变化规律．应用委托-代理理论推导出资本市场的最优激励契约．分析经纪人过度自信及交易次数对最优激励契约的影响,给出了在无法观测到经纪人过度自信程度的条件下．投资者实际提供的契约与最优激励契约间的变化关系．