Recent News
Computer Science Colloquium will discuss strategies for sustainable AI data centers
March 10, 2025
Dissertation defense: Alyshia Bustos
March 5, 2025
Computer Science undergraduate honored for cybersecurity research
February 20, 2025
Associate Professor Matt Lakin wins PECASE Award
January 31, 2025
News Archives
[Colloquium] Criticality Metrics on Sorting Algorithms
March 6, 2014
Watch Colloquium:
- Date: Thursday, March 6, 2014
- Time: 11:00 am - 12:00 pm
- Place: Mechanical Engineering 218
Thomas B. Jones
PhD Candidate
University of New Mexico
Department of Computer Science
Criticality Metrics on Sorting Algorithms
Traditional computation assumes all faults must be suppressed, but now programs operate in environments where faults are inevitable. Faults can cause a computation to return incorrect results, but little is known about which faults are likely to cause the worst damage. This talk will introduce a measure called criticality to formalize that notion. I measure the degree of damage caused in a sorting algorithm’s output by faults in each comparison. The criticality score of each comparison fault is then used to aid in the discovery of points where programs can be armored against faults. I also show that an attacker can use this criticality information to plan an attack against sorting algorithms that will do more damage than a random attack.