With the continuous expansion of data availability in many large-scale, complex, and networked systems, such as surveillance, security, Internet, and finance, it becomes critical to advance the fundamental understanding of knowledge discovery and analysis from raw data to support decision-making processes. Although existing knowledge discovery and data engineering techniques have shown great success in many real-world applications, the problem of learning from imbalanced data (the imbalanced learning problem) is a relatively new challenge that has attracted growing attention from both academia and industry. The imbalanced learning problem is concerned with the performance of learning algorithms in the presence of underrepresented data and severe class distribution skews. Due to the inherent complex characteristics of imbalanced data sets, learning from such data requires new understandings, principles, algorithms, and tools to transform vast amounts of raw data efficiently into information and knowledge representation. In this paper, we provide a comprehensive review of the development of research in learning from imbalanced data. Our focus is to provide a critical review of the nature of the problem, the state-of-the-art technologies, and the current assessment metrics used to evaluate learning performance under the imbalanced learning scenario. Furthermore, in order to stimulate future research in this field, we also highlight the major opportunities and challenges, as well as potential important research directions for learning from imbalanced data.

Learning from Imbalanced Data

Computation-intensive design problems are becoming increasingly common in manufacturing industries. The computation burden is often caused by expensive analysis and simulation processes in order to reach a comparable level of accuracy as physical testing data. To address such a challenge, approximation or metamodeling techniques are often used. Metamodeling techniques have been developed from many different disciplines including statistics, mathematics, computer science, and various engineering disciplines. These metamodels are initially developed as “surrogates” of the expensive simulation process in order to improve the overall computation efficiency. They are then found to be a valuable tool to support a wide scope of activities in modern engineering design, especially design optimization. This work reviews the state-of-the-art metamodel-based techniques from a practitioner’s perspective according to the role of metamodeling in supporting design optimization, including model approximation, design space exploration, problem formulation, and solving various types of optimization problems. Challenges and future development of metamodeling in support of engineering design is also analyzed and discussed.Copyright © 2006 by ASME

https://www.sfu.ca/~gwa5/index_files/Microsoft%20Word%20-%20ASME-Metamodeling-6-12-06.pdf

Review of Metamodeling Techniques in Support of Engineering Design Optimization

Vaitoskirja pyrkii vastaamaan kahteen kysymykseen: Onko mahdollista muotoilla tuotannon teoriaa? Jos on, johtaako teoria tehokkuuden kohoamiseen, kun sita sovelletaan rakentamiseen?

Historiallinen analyysi paljastaa, etta 1900-luvulla on kehitetty ja sovellettu kolmea eri tuotannon kasitteellista mallia. Tuotanto on nahty muunnoksena, virtana ja arvontuottona. Kaikkia naita malleja voidaan perustellusti pitaa valttamattomina tuotannon ymmartamiseksi, ja niita tulisi siten soveltaa rinnakkain. Vaitostutkimuksessa esitettavaa tuotannon mallia kutsutaan tuotannon TFV-teoriaksi (Transformation, Flow, Value generation).

Tutkimuksen tulokset osoittavat, etta tehokkuutta alentavia ongelmia esiintyy yleisesti rakentamisen kaikissa vaiheissa: rakennuttajan paatoksenteossa, suunnittelun ohjauksessa ja rakennustuotannon ohjauksessa. TFV-teoria selittaa suurelta osin rakentamisen ongelmien syntymisen. Tarkasteltaessa pioneeriyritysten kehittamishankkeita, joissa on sovellettu TFV-teorian tiettyja ydinpiirteita, voidaan havaita, etta talle teorialle perustuvat menetelmat tuottavat merkittavia kustannussaastoja ja muita hyotyja. TFV-teorian avulla voidaan siten suunnata uusiin tehokkaampiin menetelmiin tahtaavaa kokeilua ja kehitystyota seka rakentamisen kaytannon etta tutkimuksen piirissa.

/pdf/an-exploration-towards-a-production-theory-and-its-1y5pohraps.pdf

An exploration towards a production theory and its application to construction

The purpose of this article is to provide a survey of the various stages in the development of response surface methodology RSM. The coverage of these stages is organized in three parts that describe the evolution of RSM since its introduction in the early 1950s. Part I covers the period, 1951-1975, during which the so-called classical RSM was developed. This includes a review of basic experimental designs for fitting linear response surface models, in addition to a description of methods for the determination of optimum operating conditions. Part II, which covers the period, 1976-1999, discusses more recent modeling techniques in RSM, in addition to a coverage of Taguchi's robust parameter design and its response surface alternative approach. Part III provides a coverage of further extensions and research directions in modern RSM. This includes discussions concerning response surface models with random effects, generalized linear models, and graphical techniques for comparing response surface designs. Copyright © 2010 John Wiley & Sons, Inc.

Response surface methodology

Probabilistic optimization design offers tools for making reliable decisions with the consideration of uncertainty associated with design variables/parameters and simulation models. In a probabilistic design, such as reliability-based design and robust design, the design feasibility is formulated probabilistically such that the probability of the constraint satisfaction (reliability) exceeds the desired limit. The reliability assessment for probabilistic constraints often involves an iterative procedure; therefore, two loops are involved in a probabilistic optimization. Due to the double-loop procedure, the computational demand is extremely high. To improve the efficiency of a probabilistic design, a novel method – sequential optimization and reliability assessment (SORA) is developed in this paper. The SORA method employs a single-loop strategy where a serial of cycles of optimization and reliability assessment is employed. In each cycle optimization and reliability assessment are decoupled from each other; no reliability assessment is required within optimization and the reliability assessment is only conducted after the optimization. The key concept of the proposed method is to shift the boundaries of violated deterministic constraints (with low reliability) to the feasible direction based on the reliability information obtained in the previous cycle. Hence the design is quickly improved from cycle to cycle and the computational efficiency is improved significantly. Two engineering applications, the reliability-based design for vehicle crashworthiness of side impact and the integrated reliability and robust design of a speed reducer, are presented to demonstrate the effectiveness of the SORA method.Copyright © 2002 by ASME

Sequential Optimization and Reliability Assessment Method for Efficient Probabilistic Design

Computer-Based Robust Engineering : Essentials for DFSS

A process involves collecting data relating to a particular condition and parsing the data from an original set of variables into subsets. For each subset defined, Mahalanobis distances are computed for known normal and abnormal values and the square root of these Mahalanobis distances is computed. A multiple Mahalanobis distance is calculated based upon the square root of Mahalanobis distances. Signal to noise ratios are obtained for each run of an orthogonal array in order to identify important subsets. This process has applications in identifying important variables or combinations thereof from a large number of potential contributors to a condition. The multidimensional system is robust and performs predictive data analysis well even when there are incidences of multi-collinearity and variables with zero standard deviations in reference group or unit space. Separate methods are provided: adjoint matrix Gram-Schmidt's method for multi-collinearity problems, and modified Gram-Schmidt method for the cases where there are variables with zero standard deviation to achieve data analysis.

Multivariate data analysis method and uses thereof

Advantages and Limitations of MTS and MTGS

Enhancing Robust Design with the Aid of TRIZ and Axiomatic Design

The Mahalanobis-Taguchi System is a diagnosis and predictive method for analyzing patterns in multivariate cases. The goal of this study is to compare the ability of the Mahalanobis-Taguchi System and a neural network to discriminate using small data sets. We examine the discriminant ability as a function of data set size using an application area where reliable data is publicly available. The study uses the Wisconsin Breast Cancer study with nine attributes and one class.Copyright © 2005 by ASME

Genichi Taguchi

Papers

Computer-Based Robust Engineering : Essentials for DFSS

Multivariate data analysis method and uses thereof

Advantages and Limitations of MTS and MTGS

Enhancing Robust Design with the Aid of TRIZ and Axiomatic Design

A Comparison Study of Mahalanobis-Taguchi System and Neural Network for Multivariate Pattern Recognition