Hybrid - Regression and Time Series Mixture Approaches to Predict Resilience

#Reliability #IEEE #Boston #Chapter #statistics, #models, #resilience
Share

Sponsor:   IEEE Boston/Providence/New Hampshire Reliability Chapter

                Please visit https://r1.ieee.org/boston-rl/

Host:        IEEE Boston/Providence/New Hampshire Reliability Chapter


Resilience engineering is the ability to build and sustain a system that can deal effectively with disruptive events. Previous resilience engineering research focuses on metrics to quantify resilience and models to characterize system performance. However, resilience metrics are normally computed after disruptions have occurred and existing models lack the ability to predict one or more shocks and subsequent recoveries.

To address these limitations, this talk presents three alternative approaches to model system resilience with statistical techniques based on (i) regression, (ii) time series, and (iii) a combination of regression and time series to track and predict how system performance will change when exposed to multiple shocks and stresses of different intensity and duration, provide structure for planning tests to assess system resilience against particular shocks and stresses and guide data collection necessary to conduct tests effectively.

These modeling approaches are general and can be applied to systems and processes in multiple domains. A historical data set on job losses during the 1980 recessions in the United States is used to assess the predictive accuracy of these approaches. Goodness-of-fit measures and confidence intervals are computed and interval-based and point-based resilience metrics are predicted to assess how well the models perform on the data set considered. The results suggest that resilience models based on statistical methods such as multiple linear regression and multivariate time series models are capable of modeling and predicting resilience curves exhibiting multiple shocks and subsequent recoveries. However, models that combine regression and time series account for changes in performance due to current and time-delayed effects from disruptions most effectively, demonstrating superior performance in long-term predictions and higher goodness-of-fit despite increased parametric complexity.



  Date and Time

  Location

  Hosts

  Registration



  • Date: 06 Mar 2024
  • Time: 05:30 PM to 07:00 PM
  • All times are (UTC-05:00) Eastern Time (US & Canada)
  • Add_To_Calendar_icon Add Event to Calendar

This Meeting is to be delivered in-person at MIT Lincoln Lab Main Cafeteria, 244 Wood St, Lexington, MA 02421, and virtually.

At registration, you must provide a valid e-mail address to receive the Webinar Session link approximately 15 hours before the event.  The link will only be sent to the e-mail address entered with your registration.  Please double-check for spelling errors.  If you haven't received the e-mail as scheduled, please check your spam folder and alternate e-mail accounts before contacting the host.

  • Lincoln Laboratory
  • 244 Wood St
  • Lexington, Massachusetts
  • United States 02421
  • Building: Main Cafeteria

  • Contact Event Hosts
  • James P. (Jay) Yakura, Chair

    IEEE Boston/Providence/New Hampshire Reliability Chapter

  • Co-sponsored by Boston Section Life Member Affinity Group
  • Starts 18 February 2024 02:00 AM
  • Ends 05 March 2024 05:30 PM
  • All times are (UTC-05:00) Eastern Time (US & Canada)
  • No Admission Charge


  Speakers

Priscila Silva Priscila Silva of University of Massachusetts- Dartmouth

Topic:

Regression and Time Series Mixture Approaches to Predict Resilience

Resilience engineering is the ability to build and sustain a system that can deal effectively with disruptive events. Previous resilience engineering research focuses on metrics to quantify resilience and models to characterize system performance. However, resilience metrics are normally computed after disruptions have occurred and existing models lack the ability to predict one or more shocks and subsequent recoveries.

To address these limitations, this talk presents three alternative approaches to model system resilience with statistical techniques based on (i) regression, (ii) time series, and (iii) a combination of regression and time series to track and predict how system performance will change when exposed to multiple shocks and stresses of different intensity and duration, provide structure for planning tests to assess system resilience against particular shocks and stresses and guide data collection necessary to conduct tests effectively.

These modeling approaches are general and can be applied to systems and processes in multiple domains. A historical data set on job losses during the 1980 recessions in the United States is used to assess the predictive accuracy of these approaches. Goodness-of-fit measures and confidence intervals are computed and interval-based and point-based resilience metrics are predicted to assess how well the models perform on the data set considered. The results suggest that resilience models based on statistical methods such as multiple linear regression and multivariate time series models are capable of modeling and predicting resilience curves exhibiting multiple shocks and subsequent recoveries. However, models that combine regression and time series account for changes in performance due to current and time-delayed effects from disruptions most effectively, demonstrating superior performance in long-term predictions and higher goodness-of-fit despite increased parametric complexity.

Biography:

Priscila Silva is a Ph.D. candidate in Electrical and Computer Engineering at the University of Massachusetts Dartmouth (UMassD). She received her MS degree in Computer Engineering from UMassD in 2022, and her BS degree in Electrical Engineering from Federal University of Ouro Preto (UFOP) in 2017, In Brazil. She works under the supervision of Dr. Lance Fiondella in the dependable software and system lab at UMassD, where they have projects supported by the United States Military Academy West Point, Air Force, and NSF.

She has published three (3) peer-reviewed first-author conference papers with an additional four (4) first-author journal articles under review or in preparation. She is co-author of six (6) additional published conference papers, with another three (3) journal articles under review. Her very first peer-reviewed paper on which she served as the first author was published in the proceedings of the 2022 Annual Symposium on Reliability and Maintainability (RAMS), receiving Second Place in the Thomas L. Fagan Jr., RAMS Student Paper Award Competition.

Her research interests include system reliability and resilience engineering for performance evaluation, including computer, cyber-physical, infrastructure, finance, and environment domains. For her Ph.D. dissertation, she has been working on statistical modeling techniques to predict system recovery time after disruptive events, which will enable test planning and assessment to support emergency management teams to optimally allocate resources to restorative activities.





Agenda

5:30 PM     Light repast and Networking

6:00  PM   Technical Presentation

6:45 PM    Questions and Answers

7:00 PM    Adjournment



The meeting is open to all.  You do not need to belong to the IEEE to attend this event; however, we welcome your consideration of IEEE membership as a career enhancing technical affiliation.

There is no cost to register or attend, but registration is required.