Learning-Based Preference Modeling in Engineering Design Decision-Making

1999 ◽  
Vol 123 (2) ◽  
pp. 191-198 ◽  
Author(s):  
Jie Wan ◽  
Sundar Krishnamurty
Keyword(s):  
Engineering Design ◽  
Deductive Reasoning ◽  
Integrated Approach ◽  
Design Decision ◽  
Learning Method ◽  
Preference Modeling ◽  
Modeling Process ◽  
Preference Representation ◽  
Elicitation Process ◽  
Effective Integration

Focusing on the efforts towards a consistent preference representation in decision based engineering design, this paper presents a learning-based comparison and preference modeling process. Through effective integration of a deductive reasoning-based on designer’s outcome ranking in a lottery questions-based elicitation process, this work offers a reliable framework for formulating utility functions that reflect designer’s priorities accurately and consistently. It is expected that this integrated approach will reduce designer’s cognitive burden, and lead to accurate and consistent preference representation. Salient features of this approach include a linear programming based dynamic preference learning method and a logical analysis of preference inconsistencies. The development of this method and its utilization in engineering design are presented in the context of a mechanism design problem and the results are discussed.

Towards a Consistent Preference Representation in Engineering Design

1998 ◽  
Author(s):  
Jie Wan ◽  
Sundar Krishnamurty
Keyword(s):  
Engineering Design ◽  
Deductive Reasoning ◽  
Utility Functions ◽  
Design Environment ◽  
Preference Information ◽  
Beam Design ◽  
Preference Representation ◽  
Complex Engineering ◽  
Local Utility Functions

Abstract Multiattribute utility theory is commonly used to define and represent the decision-maker’s preferences under conditions of uncertainty and risk. A major issue in implementing this approach deals with the identification and generation of appropriate utility functions, especially in an often nonlinear and complex engineering design environment. Typically, the decision-maker’s preferences are provided through lottery questions rather than based on deductive reasoning to reflect the nonlinear tradeoffs among the attributes. The use of such an intuitive procedure can lead to inconsistent and inexact preference information that may result in inaccuracy and rank reversal problems. Alternatively, this paper presents an Interactive Preference-Modeling (IPM) method towards a consistent preference representation in engineering design. Focusing on the preference orientation by implicitly articulating the designer’s priorities, this method provides a methodical framework to check and eliminate inconsistency in preference information, and to accurately express preferences through rational pairwise comparisons. The development of IPM method and its utilization in the determination of the system utility function from a consistent set of local utility functions are presented in the context of a beam design problem and the results are discussed.

Arrow's Theorem and Engineering Design Decision Making

1999 ◽  
Vol 11 (4) ◽  
pp. 218-228 ◽  
Author(s):  
Michael J. Scott ◽  
Erik K. Antonsson
Keyword(s):  
Decision Making ◽  
Engineering Design ◽  
Design Decision ◽  
Arrow’S Theorem ◽  
Arrow's Theorem

Generating Pseudo-Data to Enhance the Performance of Classification-Based Engineering Design: A Preliminary Investigation

2020 ◽  
Author(s):  
Xianping Du ◽  
Onur Bilgen ◽  
Hongyi Xu
Keyword(s):  
Machine Learning ◽  
Engineering Design ◽  
Real World ◽  
Surrogate Model ◽  
Preliminary Investigation ◽  
Learning Model ◽  
Classification Model ◽  
Design Decision ◽  
Large Dataset ◽  
Machine Learning Model

Abstract Machine learning for classification has been used widely in engineering design, for example, feasible domain recognition and hidden pattern discovery. Training an accurate machine learning model requires a large dataset; however, high computational or experimental costs are major issues in obtaining a large dataset for real-world problems. One possible solution is to generate a large pseudo dataset with surrogate models, which is established with a smaller set of real training data. However, it is not well understood whether the pseudo dataset can benefit the classification model by providing more information or deteriorates the machine learning performance due to the prediction errors and uncertainties introduced by the surrogate model. This paper presents a preliminary investigation towards this research question. A classification-and-regressiontree model is employed to recognize the design subspaces to support design decision-making. It is implemented on the geometric design of a vehicle energy-absorbing structure based on finite element simulations. Based on a small set of real-world data obtained by simulations, a surrogate model based on Gaussian process regression is employed to generate pseudo datasets for training. The results showed that the tree-based method could help recognize feasible design domains efficiently. Furthermore, the additional information provided by the surrogate model enhances the accuracy of classification. One important conclusion is that the accuracy of the surrogate model determines the quality of the pseudo dataset and hence, the improvements in the machine learning model.

A Framework for Sufficiency Estimation of Engineering Design Models

2001 ◽  
Author(s):  
Rajesh Radhakrishnan ◽  
Daniel A. McAdams
Keyword(s):  
Engineering Design ◽  
Utility Theory ◽  
Negative Impact ◽  
Model Construction ◽  
Design Decision ◽  
Model Accuracy ◽  
Design Models ◽  
Experimental Trials ◽  
And Behavior ◽  
Cost Factors

Abstract Engineering design models are aids that provide the designer with the ability to visualize the form and predict the nature and behavior of any product. In each stage of design, these models are used to predict the result of, or guide, design specifications, at a time when the design can still be changed with minimal negative impact. To ensure the downstream validity of these specifications or decisions, the designer must construct models that have sufficient accuracy and resolution. Determining the goodness of a model for a particular design decision or specification is a fundamental and pervasive question in engineering. Though fast to construct, and generally inexpensive, models based on estimation and approximation may not provide information of sufficient quality to make an accurate evaluation. In contrast, the crispness and depth of information gained from detailed computational analysis or experimental trials may come at too great an expense. Hence, a key aspect of model construction in design is deciding whether a model is appropriate for a particular design specification or evaluation considering accuracy and cost factors. This paper explores the application of utility theory to the model construction problem. We also discuss how estimated model accuracy affects the confidence of selecting a particular model. We present this research through application to a racecar sway bar.

A Preliminary Study of Novice and Expert Users’ Decision-Making Procedures During Visual Trade Space Exploration

2009 ◽  
Author(s):  
David Wolf ◽  
Timothy W. Simpson ◽  
Xiaolong Luke Zhang
Keyword(s):  
Decision Making ◽  
Engineering Design ◽  
Data Visualization ◽  
Space Exploration ◽  
Computational Thinking ◽  
Test Group ◽  
Design Decision ◽  
Ongoing Research ◽  
Trade Space Exploration ◽  
Insight Into

Thanks to recent advances in computing power and speed, designers can now generate a wealth of data on demand to support engineering design decision-making. Unfortunately, while the ability to generate and store new data continues to grow, methods and tools to support multi-dimensional data exploration have evolved at a much slower pace. Moreover, current methods and tools are often ill-equipped at accommodating evolving knowledge sources and expert-driven exploration that is being enabled by computational thinking. In this paper, we discuss ongoing research that seeks to transform decades-old decision-making paradigms rooted in operations research by considering how to effectively convert data into knowledge that enhances decision-making and leads to better designs. Specifically, we address decision-making within the area of trade space exploration by conducting human-computer interaction studies using multi-dimensional data visualization software that we have been developing. We first discuss a Pilot Study that was conducted to gain insight into expected differences between novice and expert decision-makers using a small test group. We then present the results of two Preliminary Experiments designed to gain insight into procedural differences in how novices and experts use multi-dimensional data visualization and exploration tools and to measure their ability to use these tools effectively when solving an engineering design problem. This work supports our goal of developing training protocols that support efficient and effective trade space exploration.

On Validating Engineering Design Decision Support Tools

2005 ◽  
Vol 13 (2) ◽  
pp. 111-122 ◽  
Author(s):  
Andrew T. Olewnik ◽  
Kemper Lewis
Keyword(s):  
Decision Support ◽  
Engineering Design ◽  
Design Decision ◽  
Decision Support Tools ◽  
Support Tools ◽  
Design Decision Support

Manufacturing Investment and Allocation in Product Line Design Decision-Making

2005 ◽  
Author(s):  
Jeremy J. Michalek ◽  
Oben Ceryan ◽  
Panos Y. Papalambros ◽  
Yoram Koren
Keyword(s):  
Decision Making ◽  
Engineering Design ◽  
Product Line ◽  
Manufacturing Cost ◽  
Design Decision ◽  
Product Line Design ◽  
Product Decision ◽  
Market Consequences ◽  
Line Design ◽  
Is Design

An important aspect of product development is design for manufacturability (DFM) analysis that aims to incorporate manufacturing requirements into early product decision-making. Existing methods in DFM seldom quantify explicitly the tradeoffs between revenues and costs generated by making design choices that may be desirable in the market but costly to manufacture. This paper builds upon previous work coordinating models for engineering design and marketing product line decision-making by incorporating quantitative models of manufacturing investment and production allocation. The result is a methodology that considers engineering design decisions quantitatively in the context of manufacturing and market consequences in order to resolve tradeoffs, not only among performance objectives, but also between market preferences and manufacturing cost.

Taxonomy for classifying engineering decision problems and support systems

1995 ◽  
Vol 9 (5) ◽  
pp. 427-438 ◽  
Author(s):  
David G. Ullman ◽  
Bruce D'Ambrosio
Keyword(s):  
Decision Support ◽  
Engineering Design ◽  
Quantitative Data ◽  
Decision Problems ◽  
Design Decision ◽  
Decision Support Tools ◽  
Design Decisions ◽  
Wide Range ◽  
Support Tools ◽  
Need Support

AbstractThe design of even the simplest product requires thousands of decisions. Yet few of these decisions are supported with methods on paper or on computers. Is this because engineering design decisions do not need support or is it because techniques have yet to be developed that are usable on a wide basis? In considering this question a wide range of decision problem characteristics need to be addressed. In engineering design some decisions are made by individuals, others by teams – some are about the product and others about the processes that support the product – some are based on complete, consistent, quantitative data and others on sparse, conflicting, qualitative discussions. To address the reasons why so little support is used and the characteristics of potentially useful decision support tools, a taxonomy of decision characteristics is proposed. This taxonomy is used to classify current techniques and to define the requirements for an ideal engineering design decision support system.

Sign in / Sign up

Export Citation Format

Share Document