A New Multidimensional Time Series Forecasting Method Based on the EOF Iteration Scheme

In this paper a new .mnultidimensional time series forecasting scheme based on the empirical orthogonal function (EOF) stepwise iteration process is introduced. The scheme is tested in a series of forecast experiments of Nino3 SST anomalies and Tahiti-Darwin SO index. The results show that the scheme is feasible and ENSO predictable.

Replacing the Annual Budget with Business Intelligence Driver-Based Forecasts

The fixed annual budget process can be a cumbersome and static process, often failing to deliver intended benefits. Typically detached from business operations and strategic planning goals, the annual budget suffers from inherent weaknesses caused by a lack of business intelligence regarding its underlying assumptions. This weakness is well documented in existing literature and there is ample evidence of improved alternatives to static corporate financial planning. One such alternative utilizes business intelligence as an essential component in the annual budget process, along with rolling forecasts as a critical tool. Utilizing business intelligence supported, driver-based rolling forecasting can align an organization’s budget process with strategic objectives and can further the operational and financial strength of an organization, as well as maximize shareholder value. In order to fully explore this topic, this article will present a review of the conventional annual budget process and the manner in which an approach that bases financial forecasts on business intelligence drivers can align operations with strategic objectives and add value to an organization. An assessment of intelligence-supported, driver-based rolling forecasting will also be presented, demonstrating an im- proved approach to the traditional annual budgeting process.

Autonomous Kernel Based Models for Short-Term Load Forecasting

The application of support vector machines to forecasting problems is becoming popular, lately. Several comparisons between neural networks trained with error backpropagation and support vector machines have shown advantage for the latter in different domains of application. However, some difficulties still deteriorate the performance of the support vector machines. The main one is related to the setting of the hyperparameters involved in their training. Techniques based on meta-heuristics have been employed to determine appropriate values for those hyperparameters. However, because of the high noneonvexity of this estimation problem, which makes the search for a good solution very hard, an approach based on Bayesian inference, called relevance vector machine, has been proposed more recently. The present paper aims at investigating the suitability of this new approach to the short-term load forecasting problem.

APPLICATION OF CASE-BASED REASONING IN FORECASTING INSECT PESTS

ＸｉａＬｕｐｉｎｇ；Ｚｈｕｌｅｉ；ＸｉａｎｇＨｏｎｇｐｉｎｇ（ＡｎｈｕｉＩｎｓｔｉｔｕｔｅｏｆＡｃｈｉｔｅｃｔｕｒｅ）ＡＰＰＬＩＣＡＴＩＯＮＯＦＣＡＳＥ－ＢＡＳＥＤＲＥＡＳＯＮＩＮＧＩＮＦＯＲＥＣＡＳＴＩＮＧＩＮＳＥＣＴＰＥＳＴＳ￥ＸｉａＬｕｐｉｎｇ...

Thermal Energy Collection Forecasting Based on Soft Computing Techniques for Solar Heat Energy Utilization System

In recent years, introduction of alternative energy sources such as solar energy is expected. Solar heat energy utilization systems are rapidly gaining acceptance as one of the best solutions to be an alternative energy source. However, thermal energy collection is influenced by solar radiation and weather conditions. In order to control a solar heat energy utilization system as accurate as possible, it requires method of solar radiation estimation. This paper proposes the forecast technique of a thermal energy collection of solar heat energy utilization system based on solar radiation forecasting at one-day-ahead 24-hour thermal energy collection by using three different NN models. The proposed technique with application of NN is trained by weather data based on tree-based model, and tested according to forecast day. Since tree-based-model classifies a meteorological data exactly, NN will train a solar radiation with smoothly. The validity of the proposed technique is confirmed by computer simulations by use of actual meteorological data.

Relative Performance Evaluation of Competing Crude Oil Prices’ Volatility Forecasting Models: A Slacks-Based Super-Efficiency DEA Model

With the increasing number of quantitative models available to forecast the volatility of crude oil prices, the assessment of the relative performance of competing models becomes a critical task. Our survey of the literature revealed that most studies tend to use several performance criteria to evaluate the performance of competing forecasting models;however, models are compared to each other using a single criterion at a time, which often leads to different rankings for different criteria—A situation where one cannot make an informed decision as to which model performs best when taking all criteria into account. In order to overcome this methodological problem, Xu and Ouenniche [1] proposed a multidimensional framework based on an input-oriented radial super-efficiency Data Envelopment Analysis (DEA) model to rank order competing forecasting models of crude oil prices’ volatility. However, their approach suffers from a number of issues. In this paper, we overcome such issues by proposing an alternative framework.

Application of the Improved Generalized Autoregressive Conditional Heteroskedast Model Based on the Autoregressive Integrated Moving Average Model in Data Analysis

This study firstly improved the Generalized Autoregressive Conditional Heteroskedast model for the issue that financial product sales data have singular information when applying this model, and the improved outlier detection method was used to detect the location of outliers, which were processed by the iterative method. Secondly, in order to describe the peak and fat tail of the financial time series, as well as the leverage effect, this work used the skewed-t Asymmetric Power Autoregressive Conditional Heteroskedasticity model based on the Autoregressive Integrated Moving Average Model to analyze the sales data. Empirical analysis showed that the model considering the skewed distribution is effective.

THE FLUX OF LAND-BASED SOURCE POLLUTANTS FROM TUMEN RIVER SYSTEM ENTERING THE SEA OF JAPAN

The influence of land based source pollutants to marine ecological environment is principally in coastal or enclosed sea waters. Flux of land based source pollutants into the sea will be effected due to social and economic development in the Tumen River basin. Pollutant type and primary pollution factor of the Tumen River in Northeast China is described by weighted coefficient method in this paper. The results indicate that the river is organic pollution type and primary pollution factor is COD. Fresh water fraction proves that the estuary is not affected by tide cycle. COD annual flux entering the Sea of Japan calculated by zero dimension model in 1993 was 90.50 ×10 3 tons. It is estimated with emission coefficient method that the COD will be 176.4 ×10 3 and 458.6 ×10 3 tons for the years of 2000 and 2010 respectively.

Technical Note on a Track-pattern-based Model for Predicting Seasonal Tropical Cyclone Activity over the Western North Pacific

Recently, the National Typhoon Center （NTC） at the Korea Meteorological Administration launched a track-pattern-based model that predicts the horizontal distribution of tropical cyclone （TC） track density from June to October. This model is the first approach to target seasonal TC track clusters covering the entire western North Pacific （WNP） basin, and may represent a milestone for seasonal TC forecasting, using a simple statistical method that can be applied at weather operation centers. In this note, we describe the procedure of the track-pattern-based model with brief technical background to provide practical information on the use and operation of the model. The model comprises three major steps. First, long-term data of WNP TC tracks reveal seven climatological track clusters. Second, the TC counts for each cluster are predicted using a hybrid statistical-dynamical method, using the seasonal prediction of large-scale environments. Third, the final forecast map of track density is constructed by merging the spatial probabilities of the seven clusters and applying necessary bias corrections. Although the model is developed to issue the seasonal forecast in mid-May, it can be applied to alternative dates and target seasons following the procedure described in this note. Work continues on establishing an automatic system for this model at the NTC.

Modeling Freight Truck Trips in Birmingham Using Tour-Based Approach

This research study explores the use of an innovative freight tour-based approach to model truck trips as an alternative to the conventional trip-based approach. The tour-based approach is more realistic as it captures the intermediate stops of each truck and reflects the implications of those stops on vehicle miles traveled (VMT). The paper describes the truck tour-based model concept, and presents the framework of a truck tour-based travel demand forecasting approach. As a case study, Global Positioning System (GPS) truck data are used to determine origin, destination, and truck stops for trucks moving within the Birmingham, Alabama region. Such information is then utilized to model truck movements within the study region as individual truck tours. The tour-based model is ran, and the resulting performance measures are contrasted to those obtained from the conventional trip-based planning model used by the Regional Planning Commission of Greater Birmingham (RPCGB). This case study demonstrates the feasibility of using a tour-based freight demand forecasting model as an alternative to the conventional 4-step process currently used to estimate truck trips in the Birmingham region. The results and lessons learned from the Birmingham case study are expected to improve truck movement modeling practices in the region and advance the accuracy of truck travel demand forecasting models at other locations in the future.

Study of Volatility Stochastic Processes in the Context of Solvency Forecasting for Sri Lankan Life Insurers

The main business of Life Insurers is Long Term contractual obligations with a typical lifetime of 20 - 40 years. Therefore, the Solvency metric is defined by the adequacy of capital to service the cash flow requirements arising from the said obligations. The main component inducing volatility in Capital is market sensitive Assets, such as Bonds and Equity. Bond and Equity prices in Sri Lanka are highly sensitive to macro-economic elements such as investor sentiment, political stability, policy environment, economic growth, fiscal stimulus, utility environment and in the case of Equity, societal sentiment on certain companies and industries. Therefore, if an entity is to accurately forecast the impact on solvency through asset valuation, the impact of macro-economic variables on asset pricing must be modelled mathematically. This paper explores mathematical, actuarial and statistical concepts such as Brownian motion, Markov Processes, Derivation and Integration as well as Probability theorems such as the Probability Density Function in determining the optimum mathematical model which depicts the accurate relationship between macro-economic variables and asset pricing.

一种近实时全球电离层数据同化和预报系统的构建与实现

电离层天气变化正成为目前空间天气预报最重要的内容之一,建立一个可靠的、精确的电离层特征参量现报和预报系统对空间科学研究及军民用无线电信息系统保障均具有重要价值。基于国际GNSS服务组织(International GNSS Service,IGS)的地基GNSS和全球电离层无线电观测站(Global Ionospheric Radio Observatory,GIRO)数字测高仪的实时数据,以国际参考电离层(International Reference Ionosphere,IRI)模型为背景模型,采用高斯-马尔可夫-限带卡尔曼滤波同化技术,结合超大规模矩阵稀疏存储与处理方法,在云计算平台上构建完成了近实时全球电离层数据同化和预报系统(near-Real-Time Global Ionospheric Data AssiMilation and forecasting system,RT-GIDAM)。该系统具备了全球电离层TEC和电子密度的近实时(延时约5 min)、较高空间(5°×2.5°)和时间分辨率(15 min)的同化和预报功能,可为空间物理研究及相关无线电系统应用提供数据支撑。

Trace Elements in Lake Baikal: Current Status, Forecast and Monitoring Problems

Assessment of the current status of Lake Baikal proved to be based on changes in natural (“preindustrial”) chemical content in basic abiotic and biological compartments of the Lake geosystem. This approach was used to evaluate background “base-line levels” of 6 major and about 50 minor and trace ele-ments in the Lake Baikal water body using a number of most reliable data re-ported within 1992-2012. In terms of environment geochemistry Baikal is one of the purest water reservoirs on the Earth. A simple mass balance model was proposed for assessing possible anthropogenic impact on Baikal water geo-chemistry. Estimations of change trends showed that only for Na+, SO42-, Cl- and Mo growth rate of their average concentrations in the Lake occurred to be 1%, 3%, 7% and 2% in every 10 years. Space-time monitoring schedules for all water body compartments of the Lake are proposed as well as similar moni-toring programs for tributaries, precipitations, bottom sediments, aquatic biota.

我国居家养老与机构养老的协同持续发展研究

基于对2000、2010、2020年全国人口普查微观数据的实证研究,发现我国近20年来机构养老的老人数占老人总数的百分比大幅提高,而且城镇的增长幅度远大于农村。基于“中国老年健康和家庭幸福调查”数据分析,我们发现机构养老的老人多为生活不能自理、80岁及以上高龄、男性、居住地城镇、教育水平较高、没有配偶和/或没有存活子女。应用ProFamy多维家庭人口预测方法及软件的家庭人口预测表明,无论低、中、高方案,我国今后几十年的机构养老需求都将大幅度增长。例如,在机构养老中方案下,2060年居住在养老机构的65—79岁较年轻老人数和80岁及以上高龄老人数分别等于2020年的3.7倍和12.2倍。因此,建议政府部门和企业管理机构尽快努力实现居家养老与机构养老协同持续发展。

基于时空注意力机制的网约车出行需求预测模型

解决网约车运营中的乘客出行需求预测问题,以降低车辆空载率、减少乘客等待时间。在考虑乘客出行需求的动态时空依赖性的基础上,提出一种基于空间数据可视化和格兰杰因果检验的乘客出行需求空间依赖性分析方法,并结合卷积神经网络和注意力机制,建立了一种基于注意力机制的时空图卷积神经网络模型来预测乘客出行需求。实例研究表明,本模型能有效捕获乘客出行需求时空依赖性的动态特征,提升模型的预测性能,具有较高的准确性和实用性。

基于牛顿-拉弗森优化算法与注意力机制优化TCN−GRU的短期电力负荷预测

为了提升短期电力负荷预测的准确率和效率,将时间卷积网络(temporal convolutional network,TCN)、门控循环单元(gated recurrent unit,GRU)模型、牛顿-拉弗森优化算法(Newton-Raphson-based optimizer,NRBO)和注意力机制(at⁃tention mechanism,Attention)结合,提出一种NRBO-TCN-GRU-Attention的负荷预测模型。在该模型中,利用NRBO算法来优化超参数,TCN模块从负荷数据中提取特征,并将提取到的特征输入GRU模块中捕获在负荷序列中的长期依赖关系。接着,利用注意力机制强化重要特征。最后通过全连接层输出预测结果。试验结果表明,所提模型在两天及一周测试集上的决定系数、平均绝对误差、平均绝对百分比误差、均方根误差四项指标均优于其他对比模型,验证了所提模型的优越性和适用性。

基于相似日聚类和PCC-VMD-SSA-KELM模型的短期光伏功率预测

由于光伏发电的随机性和不稳定性会影响功率预测的精度,提出一种基于皮尔逊相关系数(PCC)、K-均值算法(K-means)、变分模态分解(VMD)、麻雀搜索算法(SSA)、核函数极限学习机(KELM)的光伏功率短期预测模型。首先,用PCC选取主要因素作为输入;K-均值算法进行相似日聚类,将历史数据聚类为晴天、多云和雨天;其次,VMD对原始信号进行分解,充分提取集合中的输入因素信息,提高数据质量;SSA优化KELM模型的核函数参数和正则化系数解决其参数选择敏感问题;最后,将不同序列预测值叠加得到最终预测结果。仿真结果表明,所提相似日聚类下PCC-VMD-SSA-KELM模型具有较小的预测误差。

不同预见期气候指数的干旱可预测性研究

为了探索可靠的干旱过程预测方法,以江苏省为研究区域,在识别历史干旱事件的基础上,提取干旱发生前期大尺度海气相互作用因子,形成0~12个月不同预见期的气候指数,将其分为大气环流类、海温类和大气海温综合类3个类别,分别建立具有0~12个月不同预见期的偏最小二乘回归干旱预测模型,通过模型后预测验证,定量分析干旱前期不同预见期气候指数对季节干旱过程的可预测性。结果表明,0~12个月不同预见期内,干旱同期“大气环流”模型对1994年夏秋季干旱等级预测较好;预见期为0~9个月的“海温”组的模型能预测出2010~2011年冬春连旱的过程变化,且结果稳定;干旱前期9~12个月西太平洋暖池的热力状况及西太平洋副热带高压,在不同预见期的干旱预测模型中均占较大权重,且在3个月预见期的干旱预测模型中所占权重最大,可作为江苏省季节尺度气象干旱的可预测性来源。研究成果可为季节尺度的干旱预测和干旱灾害综合防范提供科学依据。

2020年北上台风不同模式的预报误差分析

2020年台风“巴威”、“美莎克”和“海神”直接北上引发了辽宁地区大范围的强降水天气过程。针对业务数值预报模式在路径、强度和降水等方面预报存在一定偏差,利用多家数值模式台风预报资料,采用评分检验和空间检验方法对北上台风的模式预报结果进行检验评估,为以后的台风气象预报工作提供参考。结果表明:各模式对台风路径的可预报时效为72 h,随着台风逐渐北上,台风的路径预报误差偏大,台风的强度预报较稳定。在降水的MODE检验结果中,ECMWF全球模式的相似度更高,CMA_MESO_3KM区域模式对降水量级预报偏大,空报较多,但两者对台风降水都有着较好的可参考性,同时台风降水预报效果与台风路径和强度预报密切相关。

30秒坐立试验联合慢性阻塞性肺疾病患者自我评估测试评分对运动性低氧的预测价值及临床应用研究

背景慢性阻塞性肺疾病(COPD)是老年患者常见呼吸系统疾病之一,活动时外周经皮血氧饱和度(SpO_(2))下降是普遍存在的现象,与疾病预后存在相关性,目前诱导运动性低氧(EID)的评估结论不一。目的探究适合社区及居家COPD患者诱导EID的方法。方法回顾性选择中日友好医院2021年1月—2023年8月收治的76例COPD稳定期患者为研究对象。收集六分钟步行测试(6MWT)中静息SpO_(2)与运动中最低SpO_(2)(ΔSpO_(2)),以ΔSpO_(2)≥4%为EID判定标准。根据有无EID分为非EID组和EID组,比较两组的肺功能指标、30秒坐立试验(30s STST)ΔSpO_(2)、COPD患者自我评估测试(CAT)评分。探究EID与30s STSTΔSpO_(2)、CAT评分的关系;采用受试者工作特征(ROC)曲线分析30s STSTΔSpO_(2)、CAT评分及二者联合对EID的预测价值。结果非EID组(n=29)与EID组(n=47)中30s STSTΔSpO_(2)、CAT评分比较,差异有统计学意义(P<0.05);30s STSTΔSpO_(2)预测EID截断值为2%,灵敏度为59.6%,特异度为82.8%,ROC曲线下面积(AUC)为0.730(95%CI=0.614~0.846,P<0.05);CAT评分预测EID截断值为13分,灵敏度为48.9%,特异度为79.3%,AUC为0.712(95%CI=0.596~0.828,P<0.05);30s STSTΔSpO_(2)联合CAT评分预测EID的截断值为0.593,灵敏度为70.2%,特异度为72.4%,AUC为0.765(95%CI=0.659~0.871,P<0.001)。结论30s STST诱发SpO_(2)≥2%、CAT评分≥13分,需警惕EID的发生可能;30s STST、CAT评分可以成为社区及居家预测稳定期COPD患者诱导EID的测评方法。