An explainable machine learning approach to improve take-off time predictions

Accurate aircraft trajectory predictions are necessary to compute exact traffic demand figures, which are crucial for an efficient and effective air traffic flow and capacity management. At present, the uncertainty of the take-off time is one of the major contributions to the loss of trajectory predictability. In the EUROCONTROL Maastricht Upper Area Control Centre, the predicted take-off time for each individual flight relies on the information received from the Enhanced Traffic Flow Management System. However, aircraft do not always take-off at the times reported by this system due to several factors, which effects and interactions are too complex to be expressed with hard-coded rules. Previous work proposed a machine learning model that, based on historical data, was able to predict the take-off time of individual flights from a set of input features that effectively captures some of these elements. The model demonstrated to reduce by 30% the take-off time prediction errors of the current system one hour before the time that flight is scheduled to depart from the parking position. This paper presents an extension of the model, which overcomes this look-ahead time constraint and allows to improve take-off time predictions as early as the initial flight plan is received. In addition, a subset of the original set of input features has been meticulously selected to facilitate the implementation of the solution in an operational air traffic flow and capacity management system, while minimising the loss of predictive power. Finally, the importance and interactions of the input features are thoroughly analysed with additive feature attribution methods.     

Performance assessment of ensemble learning systems in financial data classification

Financial data classification plays an important role in investment and banking industry with the purpose to control default risk, improve cash and select the best customers. Ensemble learning and classification systems are becoming gradually more applied to classify financial data where outputs from different classification systems are combined. The objective of this research is to assess the relative performance of existing state‐of‐the‐art ensemble learning and classification systems with applications to corporate bankruptcy prediction and credit scoring. The considered ensemble systems include AdaBoost, LogitBoost, RUSBoost, subspace, and bagging ensemble system. The experimental results from three datasets: one is composed of quantitative attributes, one encompasses qualitative data, and another one combines both quantitative and qualitative attributes. By using ten‐fold cross‐validation method, the experimental results show that AdaBoost is effective in terms of low classification error, limited complexity, and short time processing of the data. In addition, the experimental results show that ensemble classification systems outperform existing models that were recently validated on the same databases. Therefore, ensemble classification system can be employed to increase the reliability and consistency of financial data classification task.     

基于SVM的房地产财务决策风险预测及敏感性分析

本文以房地产上市公司的财务数据为参考依据,结合相关性分析,筛选并建立影响财务风险的指标体系;利用支持向量机建立财务风险分析预测模型;并使用灰色关联理论对指标进行敏感性分析,从而得到影响财务风险最敏感的指标。     

对江苏省电力负荷进行回归预测

本文主要根据2001-2011年江苏省用电量样本数据,建立了江苏省电力负荷与人均GDP、工业化以及人口数之间的多元回归预测方程,并预测了江苏省2014-2020年总用电量数据,在此基础上提出了相应的建议。     

模糊层次分析法优化BP神经网络的交通冲突预测模型

本文通过全面剖析影响交通冲突的原因,以交通流量、道路几何设计和道路环境三方面的因素建立指标层次结构体系。提出基于模糊层次分析(FAHP)法优化BP神经网络(BPNN)的预测模型,应用于交通冲突预测。     

白音查干矿井涌水量预计

矿井涌水量计算是矿床水文地质勘查中一项重要而复杂的工作,也是矿床水文地质勘查中的根本任务之一。运用大井法和廊道法对首采区、首采工作面和开切眼涌水量进行了预计,认为大井法计算的涌水量成果较为可靠,矿井最大涌水量为296.80m3/h,为今后的矿井安全生产提供了理论依据。     

The effects of handling outliers on the performance of bankruptcy prediction models

Ratio type financial indicators are the most popular explanatory variables in bankruptcy prediction models. These measures often exhibit heavily skewed distribution because of the presence of outliers. In the absence of clear definition of outliers, ad hoc approaches can be found in the literature for identifying and handling extreme values. However, it is not clear how these different approaches can affect the predictive power of models. There seems to be consensus in the literature on the necessity of handling outliers, at the same time, it is not clear how to define extreme values to be handled in order to maximize the predictive power of models. There are two possible ways to reduce the bias originating from outliers: omission and winsorization. Since the first approach has been examined previously in the literature, we turn our attention to the latter. We applied the most popular classification methodologies in this field: discriminant analysis, logistic regression, decision trees (CHAID and CART) and neural networks (multilayer perceptron). We assessed the predictive power of models in the framework of tenfold stratified crossvalidation and area under the ROC curve. We analyzed the effect of winsorization at 1, 3 and 5% and at 2 and 3 standard deviations, furthermore we discretized the range of each variable by the CHAID method and used the ordinal measures so obtained instead of the original financial ratios. We found that this latter data preprocessing approach is the most effective in the case of our dataset. In order to check the robustness of our results, we carried out the same empirical research on the publicly available Polish bankruptcy dataset from the UCI Machine Learning Repository. We obtained very similar results on both datasets, which indicates that the CHAID-based categorization of financial ratios is an effective way of handling outliers with respect to the predictive performance of bankruptcy prediction models.     

山西王家岭煤矿底板奥灰水突水危险性评价

我国是受水害最严重的国家,奥灰水突水是我国很多煤矿在安全生产中急待解决的问题。山西王家岭煤业集团公司王家岭煤矿就面临着开采9号煤层底板奥灰水突水威胁的安全问题。结合王家岭煤矿实际钻孔数据,运用突水系数法、多层叠加抗压强度比值系数法分别对9号煤层进行突水危险性预测,并进行相互对比,根据计算所得结果得出9号煤层底板易发生突水区。     

北京R&D投入与经济增长的灰色关联分析

随着科技投入的加大,R&D投入也越来越被人们所重视。R&D投入不仅对所在区域科技与经济发展具有重要作用,而且对整个国家的科技与经济发展都具有推动作用。本文运用北京2001~2011年R&D经费投入与经济增长的有关数据变量进行灰色关联分析,并根据不同结构建立了相应模型,揭示了北京R&D经费投入与经济增长的关系。     

Assessment of land use compatibility and noise pollution at Imam Khomeini International Airport

This study aimed to investigate land use planning around airports, by employing Remote Sensing (RS) and Geographic Information Systems (GIS), in conjunction with an optimization algorithm using an Integrated Noise Model (INM) software, to establish the potential effects of aircraft noise at Imam Khomeini International Airport (IKIA) in Tehran. We also checked for land use compatibility with the noise levels around IKIA and the residents' reaction to the noise. The research was carried out in three stages: a) the establishment of Strategic Noise Map (SNM) scenarios of the airport operation in the years 2011, 2020 and 2030 using the INM software; b) the assessment of the results with emphasis on the study area land uses and application of RS and GIS and the exposure of residents at different levels of environmental noise; and c) the assessment of the intensity of aircraft noise annoyance at various times of day and night. The results indicated that developing IKIA together with the residential development will increase airport noise. Hence proper management and control of noise at IKIA is essential.