Showing 1–6 of 6 results for author: Ricketts, D

Sketching AI Concepts with Capabilities and Examples: AI Innovation in the Intensive Care Unit

Authors: Nur Yildirim, Susanna Zlotnikov, Deniz Sayar, Jeremy M. Kahn, Leigh A. Bukowski, Sher Shah Amin, Kathryn A. Riman, Billie S. Davis, John S. Minturn, Andrew J. King, Dan Ricketts, Lu Tang, Venkatesh Sivaraman, Adam Perer, Sarah M. Preum, James McCann, John Zimmerman

Abstract: Advances in artificial intelligence (AI) have enabled unprecedented capabilities, yet innovation teams struggle when envisioning AI concepts. Data science teams think of innovations users do not want, while domain experts think of innovations that cannot be built. A lack of effective ideation seems to be a breakdown point. How might multidisciplinary teams identify buildable and desirable use case… ▽ More Advances in artificial intelligence (AI) have enabled unprecedented capabilities, yet innovation teams struggle when envisioning AI concepts. Data science teams think of innovations users do not want, while domain experts think of innovations that cannot be built. A lack of effective ideation seems to be a breakdown point. How might multidisciplinary teams identify buildable and desirable use cases? This paper presents a first hand account of ideating AI concepts to improve critical care medicine. As a team of data scientists, clinicians, and HCI researchers, we conducted a series of design workshops to explore more effective approaches to AI concept ideation and problem formulation. We detail our process, the challenges we encountered, and practices and artifacts that proved effective. We discuss the research implications for improved collaboration and stakeholder engagement, and discuss the role HCI might play in reducing the high failure rate experienced in AI innovation. △ Less

Submitted 20 February, 2024; originally announced February 2024.

Comments: to appear at CHI 2024

Acute kidney injury prediction for non-critical care patients: a retrospective external and internal validation study

Authors: Esra Adiyeke, Yuanfang Ren, Benjamin Shickel, Matthew M. Ruppert, Ziyuan Guan, Sandra L. Kane-Gill, Raghavan Murugan, Nabihah Amatullah, Britney A. Stottlemyer, Tiffany L. Tran, Dan Ricketts, Christopher M Horvat, Parisa Rashidi, Azra Bihorac, Tezcan Ozrazgat-Baslanti

Abstract: Background: Acute kidney injury (AKI), the decline of kidney excretory function, occurs in up to 18% of hospitalized admissions. Progression of AKI may lead to irreversible kidney damage. Methods: This retrospective cohort study includes adult patients admitted to a non-intensive care unit at the University of Pittsburgh Medical Center (UPMC) (n = 46,815) and University of Florida Health (UFH) (n… ▽ More Background: Acute kidney injury (AKI), the decline of kidney excretory function, occurs in up to 18% of hospitalized admissions. Progression of AKI may lead to irreversible kidney damage. Methods: This retrospective cohort study includes adult patients admitted to a non-intensive care unit at the University of Pittsburgh Medical Center (UPMC) (n = 46,815) and University of Florida Health (UFH) (n = 127,202). We developed and compared deep learning and conventional machine learning models to predict progression to Stage 2 or higher AKI within the next 48 hours. We trained local models for each site (UFH Model trained on UFH, UPMC Model trained on UPMC) and a separate model with a development cohort of patients from both sites (UFH-UPMC Model). We internally and externally validated the models on each site and performed subgroup analyses across sex and race. Results: Stage 2 or higher AKI occurred in 3% (n=3,257) and 8% (n=2,296) of UFH and UPMC patients, respectively. Area under the receiver operating curve values (AUROC) for the UFH test cohort ranged between 0.77 (UPMC Model) and 0.81 (UFH Model), while AUROC values ranged between 0.79 (UFH Model) and 0.83 (UPMC Model) for the UPMC test cohort. UFH-UPMC Model achieved an AUROC of 0.81 (95% confidence interval [CI] [0.80, 0.83]) for UFH and 0.82 (95% CI [0.81,0.84]) for UPMC test cohorts; an area under the precision recall curve values (AUPRC) of 0.6 (95% CI, [0.05, 0.06]) for UFH and 0.13 (95% CI, [0.11,0.15]) for UPMC test cohorts. Kinetic estimated glomerular filtration rate, nephrotoxic drug burden and blood urea nitrogen remained the top three features with the highest influence across the models and health centers. Conclusion: Locally developed models displayed marginally reduced discrimination when tested on another institution, while the top set of influencing features remained the same across the models and sites. △ Less

Submitted 6 February, 2024; originally announced February 2024.

The TLA+ Toolbox

Authors: Markus Alexander Kuppe, Leslie Lamport, Daniel Ricketts

Abstract: We discuss the workflows supported by the TLA+ Toolbox to write and verify specifications. We focus on features that are useful in industry because its users are primarily engineers. Two features are novel in the scope of formal IDEs: CloudTLC connects the Toolbox with cloud computing to scale up model checking. A Profiler helps to debug inefficient expressions and to pinpoint the source of state… ▽ More We discuss the workflows supported by the TLA+ Toolbox to write and verify specifications. We focus on features that are useful in industry because its users are primarily engineers. Two features are novel in the scope of formal IDEs: CloudTLC connects the Toolbox with cloud computing to scale up model checking. A Profiler helps to debug inefficient expressions and to pinpoint the source of state space explosion. For those who wish to contribute to the Toolbox or learn from its flaws, we present its technical architecture. △ Less

Submitted 23 December, 2019; originally announced December 2019.

Comments: In Proceedings F-IDE 2019, arXiv:1912.09611

Journal ref: EPTCS 310, 2019, pp. 50-62

TLA+ Proofs

Authors: Denis Cousineau, Damien Doligez, Leslie Lamport, Stephan Merz, Daniel Ricketts, Hernán Vanzetto

Abstract: TLA+ is a specification language based on standard set theory and temporal logic that has constructs for hierarchical proofs. We describe how to write TLA+ proofs and check them with TLAPS, the TLA+ Proof System. We use Peterson's mutual exclusion algorithm as a simple example to describe the features of TLAPS and show how it and the Toolbox (an IDE for TLA+) help users to manage large, complex pr… ▽ More TLA+ is a specification language based on standard set theory and temporal logic that has constructs for hierarchical proofs. We describe how to write TLA+ proofs and check them with TLAPS, the TLA+ Proof System. We use Peterson's mutual exclusion algorithm as a simple example to describe the features of TLAPS and show how it and the Toolbox (an IDE for TLA+) help users to manage large, complex proofs. △ Less

Submitted 29 August, 2012; originally announced August 2012.

Comments: A shorter version of this article appeared in the proceedings of the conference Formal Methods 2012 (FM 2012, Paris, France, Springer LNCS 7436, pp. 147-154)

ACM Class: D.2.4; F.3.1

Experimental study of the impact of historical information in human coordination

Authors: Manuel Cebrian, Ramamohan Paturi, Daniel Ricketts

Abstract: We perform laboratory experiments to elucidate the role of historical information in games involving human coordination. Our approach follows prior work studying human network coordination using the task of graph coloring. We first motivate this research by showing empirical evidence that the resolution of coloring conflicts is dependent upon the recent local history of that conflict. We also cond… ▽ More We perform laboratory experiments to elucidate the role of historical information in games involving human coordination. Our approach follows prior work studying human network coordination using the task of graph coloring. We first motivate this research by showing empirical evidence that the resolution of coloring conflicts is dependent upon the recent local history of that conflict. We also conduct two tailored experiments to manipulate the game history that can be used by humans in order to determine (i) whether humans use historical information, and (ii) whether they use it effectively. In the first variant, during the course of each coloring task, the network positions of the subjects were periodically swapped while maintaining the global coloring state of the network. In the second variant, participants completed a series of 2-coloring tasks, some of which were restarts from checkpoints of previous tasks. Thus, the participants restarted the coloring task from a point in the middle of a previous task without knowledge of the history that led to that point. We report on the game dynamics and average completion times for the diverse graph topologies used in the swap and restart experiments. △ Less

Submitted 12 February, 2012; originally announced February 2012.

Comments: Presented at the Workshop of Information in Networks (WIN), New York, USA, September 2009

Maximizing profit using recommender systems

Authors: Aparna Das, Claire Mathieu, Daniel Ricketts

Abstract: Traditional recommendation systems make recommendations based solely on the customer's past purchases, product ratings and demographic data without considering the profitability the items being recommended. In this work we study the question of how a vendor can directly incorporate the profitability of items into its recommender so as to maximize its expected profit while still providing accurat… ▽ More Traditional recommendation systems make recommendations based solely on the customer's past purchases, product ratings and demographic data without considering the profitability the items being recommended. In this work we study the question of how a vendor can directly incorporate the profitability of items into its recommender so as to maximize its expected profit while still providing accurate recommendations. Our approach uses the output of any traditional recommender system and adjust them according to item profitabilities. Our approach is parameterized so the vendor can control how much the recommendation incorporating profits can deviate from the traditional recommendation. We study our approach under two settings and show that it achieves approximately 22% more profit than traditional recommendations. △ Less

Submitted 25 August, 2009; originally announced August 2009.