Network Analysis of Design Automation Literature

2017 ◽  
Author(s):  
Tinghao Guo ◽  
Jiarui Xu ◽  
Yue Sun ◽  
Yilin Dong ◽  
Neal E. Davis ◽  
...  
Keyword(s):  
Engineering Design ◽  
Design Automation ◽  
Design Research ◽  
Learning Algorithm ◽  
Small World ◽  
Citation Network ◽  
Research Literature ◽  
Rule Mining ◽  
Data Set ◽  
Topic Distribution

In this paper we present a study of citation and co-authorship networks for articles from the ASME Design Automation Conference (DAC) during the years 2002–2015. We identify key authors, show that the co-authorship network exhibits the small world network property, and reveal other insights from network structure. Results from two topic modeling methods are presented. A frequency-based model was developed to explore DAC topic distribution and evolution. Citation analysis was also conducted for each core topic. A correlation matrix and association rule mining were used to discover topic relations and to gain insights for research gaps and recommendations. A recently developed unsupervised learning algorithm, propagation mergence (PM), was applied to the DAC citation network. Influential papers and major clusters were identified and visualizations are presented. The resulting insights may be beneficial to the engineering design research community, especially with respect to determining future directions and possible actions for improvement. The data set used here is limited. Expanding to include additional relevant conference proceedings and journal articles in the future would offer a more complete understanding of the engineering design research literature.

Network Analysis of Design Automation Literature

2018 ◽  
Vol 140 (10) ◽  
Author(s):  
Tinghao Guo ◽  
Jiarui Xu ◽  
Yue Sun ◽  
Yilin Dong ◽  
Neal Davis ◽  
...  
Keyword(s):  
Engineering Design ◽  
Design Automation ◽  
Design Research ◽  
Learning Algorithm ◽  
Small World ◽  
Citation Network ◽  
Research Literature ◽  
Future Research ◽  
Data Set ◽  
Topic Distribution

In this paper, we present the results of a study of citation and co-authorship networks for articles published at the ASME Design Automation Conference (DAC) during the years 2002–2015. Two topic-modeling methods are presented for studying the DAC literature: A frequency-based model was developed to explore DAC topic distribution and evolution, as well as citation analysis for each core topic. Correlation analysis and association-rule mining were used to discover relationships between topics. A new unsupervised learning algorithm, propagation mergence (PM), was created to address identified shortcomings of existing methods and applied to study the existing DAC citation network. Influential articles and important article clusters were identified and effective visualizations created. We also investigated the DAC co-authorship network by identifying key authors and showing that the network structure exhibits small-world-network properties. The resulting insights, obtained by the both the proposed and existing methods, may be beneficial to the engineering design research community, especially with respect to determining future research directions and possible actions for improvement. The data set used here is limited; expanding to include additional relevant conference proceedings and journal articles in the future would offer a more complete understanding of the engineering design research literature.

scholarly journals Experiential Learning vs Systematic Prescriptions in Engineering Design: A Crossroads for Education

2018 ◽  
Author(s):  
G. R. Gressfc ◽  
S. Li ◽  
R. W. Brennan
Keyword(s):  
Engineering Design ◽  
Design Methodology ◽  
Design Research ◽  
Great Difficulty ◽  
Research Literature ◽  
Classical Design ◽  
Classical Methodology ◽  
Successful Design ◽  
Research And Education ◽  
Visual Domain

The systematic, non-experiential prescriptions of classical design methodology continue to have a strong presence in large segments of design research and education while another segment sees domain experience and consequent intuition and creativity as being key to successful design. In this paper the two approaches are outlined and the empirical research literature in human behaviour is employed to discern discrepancies and potential weaknesses. Results show that gaining experience in a domain intrinsically changes how one designs, which the classical methodology does not account for. For example, only designers with tactile and visual domain experience can abstract functions per the dictates of the classical (non-experiential) methodology, which means that they cannot have used the methodology to learn basic design in the first place – or did so only with great difficulty. This and other conflicts pose problems for the education of engineering design students, and to fathom their extent this paper surveys engineering design textbooks offered in Canada and the U. S.; all of the books are found to embrace the classical methodology. If they are to remain involved in preparing students for entry into industry then some aspects of their contained classical methodology must be supplanted by experiential approaches to design educatio

Roles of externalisation activities in the design process

2016 ◽  
Vol 11 (1) ◽  
pp. 34
Author(s):  
Maral Babapour Chafi
Keyword(s):  
Design Process ◽  
Design Research ◽  
Research Literature ◽  
Physical Models ◽  
Future Research ◽  
Knowledge Gaps ◽  
Design Processes ◽  
Digital Modelling

Designers engage in various activities, dealing with different materials and media to externalise and represent their form ideas. This paper presents a review of design research literature regarding externalisation activities in design process: sketching, building physical models and digital modelling. The aim has been to review research on the roles of media and representations in design processes, and highlight knowledge gaps and questions for future research.

scholarly journals Information-Theoretic Generalization Bounds for Meta-Learning and Applications

Entropy ◽  
2021 ◽  
Vol 23 (1) ◽  
pp. 126
Author(s):  
Sharu Theresa Jose ◽  
Osvaldo Simeone
Keyword(s):  
Learning Algorithm ◽  
Broad Class ◽  
Performance Measure ◽  
Training Data ◽  
Learning To Learn ◽  
Data Set ◽  
Information Theoretic ◽  
Meta Learning ◽  
Task Training ◽  
Test Sets

Meta-learning, or “learning to learn”, refers to techniques that infer an inductive bias from data corresponding to multiple related tasks with the goal of improving the sample efficiency for new, previously unobserved, tasks. A key performance measure for meta-learning is the meta-generalization gap, that is, the difference between the average loss measured on the meta-training data and on a new, randomly selected task. This paper presents novel information-theoretic upper bounds on the meta-generalization gap. Two broad classes of meta-learning algorithms are considered that use either separate within-task training and test sets, like model agnostic meta-learning (MAML), or joint within-task training and test sets, like reptile. Extending the existing work for conventional learning, an upper bound on the meta-generalization gap is derived for the former class that depends on the mutual information (MI) between the output of the meta-learning algorithm and its input meta-training data. For the latter, the derived bound includes an additional MI between the output of the per-task learning procedure and corresponding data set to capture within-task uncertainty. Tighter bounds are then developed for the two classes via novel individual task MI (ITMI) bounds. Applications of the derived bounds are finally discussed, including a broad class of noisy iterative algorithms for meta-learning.

Event detection of different English data sources based on transfer learning

2021 ◽  
pp. 1-11
Author(s):  
Yanan Huang ◽  
Yuji Miao ◽  
Zhenjing Da
Keyword(s):  
Transfer Learning ◽  
Event Detection ◽  
Visual Analysis ◽  
Learning Algorithm ◽  
Data Sources ◽  
Data Set ◽  
Data Source ◽  
Single Data Source ◽  
The Difference ◽  
Single Data

The methods of multi-modal English event detection under a single data source and isomorphic event detection of different English data sources based on transfer learning still need to be improved. In order to improve the efficiency of English and data source time detection, based on the transfer learning algorithm, this paper proposes multi-modal event detection under a single data source and isomorphic event detection based on transfer learning for different data sources. Moreover, by stacking multiple classification models, this paper makes each feature merge with each other, and conducts confrontation training through the difference between the two classifiers to further make the distribution of different source data similar. In addition, in order to verify the algorithm proposed in this paper, a multi-source English event detection data set is collected through a data collection method. Finally, this paper uses the data set to verify the method proposed in this paper and compare it with the current most mainstream transfer learning methods. Through experimental analysis, convergence analysis, visual analysis and parameter evaluation, the effectiveness of the algorithm proposed in this paper is demonstrated.

Classification of Iteration in Engineering Design Processes

1998 ◽  
Author(s):  
Michael J. Safoutin ◽  
Robert P. Smith
Keyword(s):  
Engineering Design ◽  
Design Process ◽  
Design Automation ◽  
Automated Design ◽  
Design Processes ◽  
Standard Definition ◽  
Design Iteration ◽  
Formal Study ◽  
Poor Understanding

Abstract As engineering design is subjected to increasingly formal study, an informal attitude continues to surround the topic of iteration. Today there is no standard definition or typology of iteration, no grounding theory, few metrics, and a poor understanding of its role in the design process. Existing literature provides little guidance in investigating issues of design that might be best approached in terms of iteration. We review contributions of existing literature toward the understanding of iteration in design, develop a classification of design iteration, compare iterative aspects of human and automated design, and draw some conclusions concerning management of iteration and approaches to design automation.

Impact of cancer- and patient-level factors on provider risk in the Oncology Care Model.

2021 ◽  
Vol 39 (15_suppl) ◽  
pp. e18852-e18852
Author(s):  
Basit Iqbal Chaudhry ◽  
Andrew Yue ◽  
Shuchita Kaila ◽  
Kay Sadik ◽  
Lisa Tran ◽  
...  
Keyword(s):  
Domain Knowledge ◽  
Research Literature ◽  
Risk Exposure ◽  
Progression Rate ◽  
Cell Transplant ◽  
Care Model ◽  
Data Set ◽  
Patient Level ◽  
Oncology Care ◽  
The Impact

e18852 Background: Transferring financial risk from payers to providers to align incentives is central to value-based payment (VBP) reform, including Medicare’s Oncology Care Model (OCM). We simulated the impact of selected cancer- and patient-level factors on providers’ risk in OCM for multiple myeloma (MM), due to its clinical complexity. We hypothesize that risk exposure is sensitive to factors extrinsic to the OCM methodology, including clinical phenotype, disease state and progression rate. Methods: Simulation was used to address omitted variable bias in payer data. We developed 9 key clinical MM scenarios to examine provider risk, based on conceptual frameworks that included patient- and cancer-level factors. The model was parameterized using the Medicare limited data set, research literature and domain knowledge. Twenty factors were varied for each model, e.g. age, autologous stem cell transplant (ASCT). Results: Simulations results showed MM risk for providers depended highly on cancer and patient level factors (see table). For example, high-risk patients were on average $21.5K over target while undergoing ASCT (despite risk adjustment for ASCT) and $18-28K under target for follow on maintenance (maint.) episodes. Conclusions: Provider exposure to risk in OCM is highly sensitive to factors at the cancer and patient level. The distribution of clinical phenotypes, state of disease, and rate of disease progression can significantly impact risk exposure for providers in OCM. New methodologies that model risk in more clinically granular ways are needed to improve VBP in oncology. [Table: see text]

Sign in / Sign up

Export Citation Format

Share Document