Design and implementation of the fuzzy expert system in Monte Carlo methods for fuzzy linear regression
طراحی و اجرای سیستم خبره فازی در روش های مونت کارلو برای رگرسیون خطی فازی-2019
In this study, fuzzy expert system (FES) in Monte Carlo (MC) method, which is used for estimating fuzzy linear regression model (FLRM) parameters, is applied to determine the parameter intervals, for the first time in the literature. MC method in estimating FLRM parameters is a new field of study that is very useful and time saving. However a major problem might occur in determining the parameter intervals from which the regression model parameters are supposed to come. If the intervals are calculated too large, FLRM error will be very large. Accordingly, the actual model parameters will not be obtained if the intervals are calculated too narrow. This drawback has not been addressed in the literature before and only optimization methods have been applied to achieve the best interval values. In this article, the FES is used for the first time in order to solve the problem in parameter estimation process for the FLRM in the field of statistics. For this purpose, the difference between the fuzzy observation value and fuzzy estimation value’s support set (W) is taken into account. The most appropriate intervals calculated for the parameters are those that make W as small as possible. Thus, FES is designed to determine the best intervals for the model parameters. The system knowledge base is composed of 7 fuzzy rules. As a result, it is deduced that the FLRM parameter estimates obtained from the MC method using FES are very close to the real values. The real impact of this paper will be in showing the applicability of FESs in order to solve problems that we encounter in the field of statistics by the help of linguistic expressions. Moreover, these outcomes will be useful for enriching the studies that have already focused on FLRMs and will encourage researchers to use FES to solve problems in statistics. To sum up, this study demonstrates that FESs which is used in technological devices and makes our lives easier can also be used in solving problems that we confront in the field of statistics efficiently with using linguistic expressions like human inference system.
Keywords: Fuzzy expert system | Fuzzy linear regression | Monte Carlo
Assessing environmental performance in early building design stage: An integrated parametric design and machine learning method
ارزیابی عملکرد محیطی در مرحله طراحی اولیه ساختمان: یکپارچه طراحی پارامتری و یک روش یادگیری ماشین-2019
Decisions made at early design stage have major impacts on buildings’ life-cycle environmental performance. However, when only a few parameters are determined in early design stages, the detailed design decisions may still vary significantly. This may cause same early design to have quite different environmental impacts. Moreover, default settings for unknown detailed design parameters clearly cannot cover all possible variations in impact, and Monte Carlo analysis is sometimes not applicable as parameters’ probability distributions are usually unknown. Thus, uncertainties about detailed design make it difficult for existing environmental assessment methods to support early design decisions. Thus, this study developed a quantitative method using parametric design technology and machine learning algorithms for assessing buildings’ environmental performance in early decision stages, considering uncertainty associated with detailed design decisions. The parametric design technology creates design scenarios dataset, then associated environmental performances are assessed using environmental assessment databases and building performance simulations. Based on the generated samples, a machine learning algorithm integrating fuzzy C-means clustering and extreme learning machine extracts the case-specific knowledge regarding designed buildings’ early design associated with environmental uncertainty. Proposed method is an alternative but more generally applicable method to previous approaches to assess buildings environmental uncertainty in early design stages.
Keywords: Building early design | Parametric design | Machine learning | Environmental impact | Prediction intervals
Deep learning for supervised classification of spatial epidemics
یادگیری عمیق برای طبقه بندی نظارت شده از بیماری های همه گیر فضایی-2019
In an emerging epidemic, public health officials must move quickly to contain the spread. Information obtained from statistical disease transmission models often informs the de- velopment of containment strategies. Inference procedures such as Bayesian Markov chain Monte Carlo allow researchers to estimate parameters of such models, but are computa- tionally expensive. In this work, we explore supervised statistical and machine learning methods for fast inference via supervised classification, with a focus on deep learning. We apply our methods to simulated epidemics through two populations of swine farms in Iowa, and find that the random forest performs well on the denser population, but is outperformed by a deep learning model on the sparser population.
Keywords: Deep learning | Classification | Individual-level models
Deriving reservoir operation rule based on Bayesian deep learning method considering multiple uncertainties
استخراج قانون انباره بر اساس روش یادگیری عمیق بیزی با توجه به عدم قطعیت های متعدد-2019
Reservoir operation rules play a key role in real-time reservoir operation; the main factors affecting operation decisions are the current reservoir status and the future inflows. However, the future reservoir inflows are stochastic and always contain uncertainty. To study the influence of inflow uncertainty on reservoir operation rules, this paper proposes a Bayesian Deep learning method that considers both model parameter uncertainty and inflow uncertainty. In the model, the Monte Carlo integration is used to convert the complex integrals of inflow probability into a summation form. Variational inference is employed to obtain the posterior distribution of model parameters. The proposed method is applied to a real-world application at Three Gorges Project on the Yangtze River. The uncertainty estimation results show that the influence of inflow uncertainty on reservoir operation rule is greater than model parameter uncertainty, and the decision of reservoir operation is more sensitive to the reservoir inflow during dry season than other seasons. The experimental results demonstrate that the proposed Bayesian deep learning performs better than the comparison method in term of hydropower generation and the root mean square errors. Moreover, the proposed method is more robust than the comparison method when considering the inflow uncertainty.
Keywords: Reservoir operation | Operation rule | Uncertainty analysis | Deep learning | Variational inference | Bayesian neural networks
An integrated stochastic fuzzy MCDM approach to the balanced scorecard-based service evaluation
یک رویکرد MCDM فازی تصادفی یکپارچه برای ارزیابی خدمات مبتنی بر کارت امتیازی متعادل-2019
The purpose of the study is to analyse the balanced scorecard (BSC)-based evaluation of the new service development (NSD) in Turkish banking sector. The proposed model includes fuzzy ANP (FANP), Monte Carlo Simulation, fuzzy TOPSIS (FTOPSIS), and fuzzy VIKOR (FVIKOR) respectively. FANP has been used for weighting the criteria, Monte Carlo Simulation has been applied to provide the stochastic values of BSC-based dimensions of NSD in banking sector. FTOPSIS and FVIKOR have been considered to rank the banks by their dimension performances. The novelty of the study is to provide an integrated model including FANP, FTOPSIS, FVIKOR, and Monte Carlo Simulation respectively. Additionally, BSC-based analysis of NSD has been applied for evaluating Turkish banking sector. The results demonstrate that the comparative analysis is coherent for ranking the alternatives and the stochastic values facilitate to obtain the immense expert evaluations under the fuzzy environment. It is identified that the performance of the foreign banks is lower than private and state banks. Hence, it can be said that especially foreign banks should develop new services to attract the attention of their customers. Within this framework, customer expectations should be defined by conducting a detailed analysis. As a result, it can be possible to increase comparative advantage in comparison with the other banks.
Keywords: Balanced scorecard | New service development | Turkey | Banking | Monte Carlo simulation | Fuzzy sets
A new macro stress testing approach for financial realignment in the Eurozone
یک روش جدید تست استرس کلان برای تجدید مالی در منطقه یورو-2019
Contrary to the common approach of stress-testing under which banks are evaluated whether they are distressed, this empirical study chooses to move from the micro stress test approach to a wider new macro stress test category. By being able to stress testing the entire economy of the Eurozone, it will permit big banks to fail and, at the same time, will open room for new banking players to enter the sector, promoting the essence of a healthy destruction. The analysis performs a battery of stress tests, by implementing VaR, Cornish-Fisher VaR, Monte Carlo VaR, Expected Shortfall, Cornish-Fisher Expected Shortfall, and Monte Carlo Expected Shortfall. At the same time, it explicitly considers the new regulatory approach of IFRS9 to incorporate extreme values from forecasted series in the distributions. The analysis also performs two versions of stress tests, one including TARGET2 and one without it. The results document that future stress tests should include TARGET2 values in order to capture a better picture of the stressed economy. The findings from these stress tests clearly illustrate that although there has been a trough after the distress call of 2008, this trough ended. These are results derived without including the TARGET2 transfers. By including the TARGET2 transfers we receive a different picture that possibly acts as a protective mechanism against any future crisis. Ca Keywords: Macroprudential policy | Financial stability | Macro stress test | Systemic risk | European Banking Union
A deep learning solution approach for high-dimensional random differential equations
یک روش راه حل یادگیری عمیق برای معادلات دیفرانسیل تصادفی با ابعاد بالا-2019
Developing efficient numerical algorithms for the solution of high dimensional random Partial Differential Equations (PDEs) has been a challenging task due to the well-known curse of dimensionality. We present a new solution approach for these problems based on deep learning. This approach is intrusive, entirely unsupervised, and mesh-free. Specifically, the random PDE is approximated by a feed-forward fully-connected deep residual network, with either strong or weak enforcement of initial and boundary constraints. Parameters of the approximating deep neural network are determined iteratively using variants of the Stochastic Gradient Descent (SGD) algorithm. The satisfactory accuracy of the proposed approach is numerically demonstrated on diffusion and heat conduction problems, in comparison with the converged Monte Carlo-based finite element results.
Keywords: Deep learning | Deep neural networks | Residual networks | Random differential equations | Curse of dimensionality | Least squares
Modeling Diversification and Spillovers of Loan Portfolios’ Losses by LHP Approximation and Copula
تنوع مدل سازی و سرریز از تلفات پرتفوی وام توسط تقریب و کوپلای LHP-2019
This paper suggests a top-down method for aggregating the economic capital of an entire banking system and decomposing it into loan sectors according to their risk contributions. We model the individual losses of loan sectors by large homogeneous portfolio (LHP) approximation based on multi-factor skew normal credit worthiness and combine them by applying static and dynamic copulas to reflect diversification effects and spillovers across loan sectors. Our method is more efficient and practically useful than typical multi-factor models using numerical integration due to the latency of risk factors in that losses are directly generated by Monte Carlo simulation using copula without knowing any risk factors. As a result of our empirical study on charge-off rates of the U.S. commercial banking system, we find that the residential real estate loan sector is the most affecting as its default risk spills over to the rest of the banking system, and hence its risk contribution to the entire banking system is large. However, the commercial real estate loan and business loan sectors are revealed to be affected sectors whose risk contributions are large, but the losses are mainly due not only to their large exposure size, but also to default contagion from others. The risk contributions of credit cards and other consumer loans as default risk affecting sectors become larger in terms of the recent conditional dependence. Lastly, using time-varying correlation analysis, we find that the subprime mortgage crisis is a systemic event that affects the entirebanking- system, while the commercial real estate and the dotcom bubble crises are sector-wide systemic events.
Keywords: Multi-factor mode | Copula | Loss distribution | Diversification | Spillover
Measuring management practices
سنجش اقدامات مدیریتی-2018
Good management practices are remarkably difficult to robustly measure, especially when unique data on firms and their managers are not available. We propose a new model estimated with Bayesian techniques that requires only the usual accounting data on inputs and outputs and thus can be applied to any firm. We show that our management practices estimates are more than 90% correlated with existing state-of-the-art measures from a very specialized data set by Bloom and Van Reenen (2007). We also obtain very high correlations when conducting an extensive Monte Carlo analysis. Finally, we show that frontier-based methods previously used to estimate management practices do not provide good approximations.
keywords: Management practices |Productivity |Cost efficiency |Bayesian methods
آزمایش جامعیت جداسازی انتروپیکی در سطوح پلیمری
سال انتشار: 2018 - تعداد صفحات فایل pdf انگلیسی: 6 - تعداد صفحات فایل doc فارسی: 24
این مقاله به مطالعه اثرات جداسازی انتروپیکی برروی سطوح مذاب یکپارچه و بدون یکپارچه ای تحقیق میکند. برای مواد مذاب یکپارچه، ما برروی جداسازی انتهای زنجیره به سطح تمرکز کرده، و برای مواد مذاب غیر یکپارچه، جداسازی پلیمرهای کوتاه از سطح را می آزماییم. اشکال جهانی برای پروفیل های غلظت آنها پیش بینی شده است، اما این مشتقات بر تقریب میانگین تکیه می کنند، که تنها به صورت تقریبی رفتارهای منفرد حجم را بررسی می کند. برای تست اینکه آیا پیش بینی ها زمانی که پلیمرها به طور دقیق مانع از همپوشانی شدن می شوند یا خیر، ما محاسبات میانگین را با شبیه سازی مونت کارلو که بر اساس همان مدل مشابه انجام شده است، مقایسه میکنیم. به استثنای افزایش قابل توجهی در امتداد بخش آماری، اجرای دقیق تعاملات غیرمتمرکز اثر نسبتا کمی بر پیش بینی های میانگین دارد. به طور خاص، پروفیل های جهانی پیش بینی شده توسط تئوری میانگین دقیق هستند.
|مقاله ترجمه شده|