More Like this

1. Verification & Validation of Computational Models associated with the Mechanics of Materials with the Mechanics of Materials.

   https://doi.org/www.dx.doi.org/10.7449/VandV_1.  

   The Minerals Metals & Materials Society, Inc (April 2019, Verification & Validation of Computational Models associated with the Mechanics of Materials with the Mechanics of Materials.)

   Predictive computational models associated with the mechanics of materials (MOM) offer great potential for enabling large reductions in the cost and time to develop new products and manufacturing procedures. Unfortunately, this potential is currently limited because very rarely are such models adequately and broadly proven to yield trustworthy, accurate, quantitative results for which the level of uncertainty has been quantified. In this regard, the need for rigorous verification and validation (V&V) of these models cannot be overestimated, yet is extremely lacking within the relevant MOM communities. There is thus a strong need to help these communities accelerate the widespread adoption and implementation of such V&V activities. In this vein, concise definitions of verification and validation have been provided by the American Society of Mechanical Engineers (ASME),1 and can be applied here as well: • Verification: The process of determining that a computational model accurately represents the underlying mathematical model and its solution • Validation: The process of determining the degree to which a model is an accurate representation of the real world from the perspective of the intended uses of the model The overarching goal of this workshop and report is thus to help facilitate the widespread and rigorous adoption of V&V by both computational modelers and experimentalists in MOM-related communities. 

   more » « less
2. Regression with Comparisons: Escaping the Curse of Dimensionality with Ordinal Information

   Xu, Yichong; Balakrishnan, Sivaraman; Dubrawski, Arthur; Singh, Aarti (January 2020, Journal of machine learning research)

   null (Ed.)
3. Elusive Nonsolvated Cocrystals of Aspirin: Two Polymorphs with Bipyridine Discovered with the Assistance of Mechanochemistry

   https://doi.org/10.1021/acs.cgd.8b00104  

   Oburn, Shalisa M.; Ray, Olivia A.; MacGillivray, Leonard R. (February 2018, Crystal Growth & Design)
4. Designing Physical Representations of STEM Concepts With College Students With Cognitive Impairments

   Parmar, Ruchita; Bruce, Carrie M. (December 2020, CMD-IT/ACM Richard Tapia Celebration of Diversity in Computing Conference)

   null (Ed.)

   The Georgia Tech Excel program's Collaborative Design course involves 12 college students with varying degrees of cognitive impairment. This course is meant to support authentic inclusive design and enable equitable access to design language and processes for these as well as other students with impairments which could range from dexterity issues to intellectual developmental disorders or executive functioning issues. This paper primarily focuses on the research that we conducted to design and run a pilot module within the course that focuses on exploring hands-on physical represenations of online physics simulations. 

   more » « less
5. Nonparametric Regression with Comparisons: Escaping the Curse of Dimensionality with Ordinal Information

   Xu, Yichong; Muthakana, Hariank; Balakrishnan, Sivaraman; Singh, Aarti; Dubrawski, Artur (July 2018, International Conference on Machine Learning)

   In supervised learning, we leverage a labeled dataset to design methods for function estimation. In many practical situations, we are able to obtain alternative feedback, possibly at a low cost. A broad goal is to understand the usefulness of, and to design algorithms to exploit, this alternative feedback. We focus on a semi-supervised setting where we obtain additional ordinal (or comparison) information for potentially unlabeled samples. We consider ordinal feedback of varying qualities where we have either a perfect ordering of the samples, a noisy ordering of the samples or noisy pairwise comparisons between the samples. We provide a precise quantification of the usefulness of these types of ordinal feedback in non-parametric regression, showing that in many cases it is possible to accurately estimate an underlying function with a very small labeled set, effectively escaping the curse of dimensionality. We develop an algorithm called Ranking-Regression (RR) and analyze its accuracy as a function of size of the labeled and unlabeled datasets and various noise parameters. We also present lower bounds, that establish fundamental limits for the task and show that RR is optimal in a variety of settings. Finally, we present experiments that show the efficacy of RR and investigate its robustness to various sources of noise and model-misspecification. 

   more » « less