dfgcompare: a library to support process variant analysis through Markov models

Abstract Background Data-driven process analysis is an important area that relies on software support. Process variant analysis is a sort of analysis technique in which analysts compare executed process variants, a.k.a. process cohorts. This comparison can help to identify insights for improving processes. There are a few software supports to enable process cohort comparison based on the frequencies of process activities and performance metrics. These metrics are effective in cohort analysis, but they cannot support cohort comparison based on the probability of transitions among states, which is an important enabler for cohort analysis in healthcare. Results This paper defines an approach to compare process cohorts using Markov models. The approach is formalized, and it is implemented as an open-source python library, named dfgcompare. This library can be used by other researchers to compare process cohorts. The implementation is also used to compare caregivers’ behavior when prescribing drugs in the Stockholm Region. The result shows that the approach enables the comparison of process cohorts in practice. Conclusions We conclude that dfgcompare supports identifying differences among process cohorts.

Download Full-text

Detecting and Localizing Anomalies in Container Clusters Using Markov Models

Electronics ◽

10.3390/electronics9010064 ◽

2020 ◽

Vol 9 (1) ◽

pp. 64 ◽

Cited By ~ 1

Author(s):

Areeg Samir ◽

Claus Pahl

Keyword(s):

Performance Metrics ◽

Markov Models ◽

System Availability ◽

Root Cause ◽

Hierarchical Hidden Markov Model ◽

Timely Fashion ◽

Time Variations ◽

And Performance ◽

Container Orchestration ◽

Detection And Localization

Detecting the location of performance anomalies in complex distributed systems is critical to ensuring the effective operation of a system, in particular, if short-lived container deployments are considered, adding challenges to anomaly detection and localization. In this paper, we present a framework for monitoring, detecting and localizing performance anomalies for container-based clusters using the hierarchical hidden Markov model (HHMM). The model aims at detecting and localizing the root cause of anomalies at runtime in order to maximize the system availability and performance. The model detects response time variations in containers and their hosting cluster nodes based on their resource utilization and tracks the root causes of variations. To evaluate the proposed framework, experiments were conducted for container orchestration, with different performance metrics being used. The results show that HHMMs are able to accurately detect and localize performance anomalies in a timely fashion.

Download Full-text

Adductor Muscle Injuries in UEFA Soccer Athletes: A Matched-Cohort Analysis of Injury Rate, Return to Play, and Player Performance From 2000 to 2015

Orthopaedic Journal of Sports Medicine ◽

10.1177/23259671211023098 ◽

2021 ◽

Vol 9 (9) ◽

pp. 232596712110230

Author(s):

Ophelie Lavoie-Gagne ◽

Nabil Mehta ◽

Sumit Patel ◽

Matthew R. Cohn ◽

Enrico Forlenza ◽

...

Keyword(s):

Muscle Injury ◽

Performance Metrics ◽

Cohort Analysis ◽

Multivariable Analysis ◽

Injury Rate ◽

Adductor Muscle ◽

Return To Play ◽

Level Of Evidence ◽

And Performance ◽

Player Performance

Background: The effects of adductor muscle injury on performance in soccer athletes are unknown. Purpose: To (1) determine the rate and time to return to play (RTP) after adductor muscle injury, (2) investigate the rate of reinjury after RTP, and (3) investigate any long-term effects of injury on elite soccer player performance. Study Design: Cohort study; Level of evidence, 3. Methods: Using publicly available records, athletes sustaining adductor muscle injury were identified across the 5 major European soccer leagues (English Premier League, Bundesliga, La Liga, Ligue 1, and Serie A) between 2000 and 2015. Injured athletes were matched to controls by demographic characteristics and performance metrics from 1 season before the index timepoint. Investigations included the rate of RTP, reinjuries, player characteristics associated with RTP within 2 seasons, player availability, field time, and performance metrics during the 4 seasons after injury. Results: A total of 671 players with adductor muscle injury were included. Based on time to RTP, 86% of injuries were mild to moderate (4-28 days missed), and 4% required surgical intervention. Players with adductor muscle injury were absent for a median of 22 days (range, 1-700 days) and 4 games (range, 1-76 games). A total of 521 (78%) players returned at the same level, with no demographic or clinical characteristics associated with RTP on the multivariable regression. Of those returning to play, 143 (21%) experienced adductor reinjury. After RTP, defenders demonstrated decreased field time compared with controls ( P < .05). As compared with controls, defenders and midfielders scored more points and goals per game during the season of the injury ( P < .01), while attackers recorded more goals and assists per game the season after injury ( P < .05). Conclusion: Only 3 in 4 players (78%) returned to participate in an official match, and the reinjury rate was high (21%). After RTP, defenders demonstrated decreased field time versus controls. On the other hand, defenders and midfielders recorded more points and goals per game, while attackers recorded more goals and assists per game versus controls. Although the multivariable analysis results did not identify player characteristics associated with RTP, there was a position-dependent association on player performance after RTP.

Download Full-text

Return to Play and Player Performance After Meniscal Tear Among Elite-Level European Soccer Players: A Matched Cohort Analysis of Injuries From 2006 to 2016

Orthopaedic Journal of Sports Medicine ◽

10.1177/23259671211059541 ◽

2022 ◽

Vol 10 (1) ◽

pp. 232596712110595

Author(s):

Ophelie Z. Lavoie-Gagne ◽

Avinaash Korrapati ◽

Julia Retzky ◽

David N. Bernstein ◽

Connor C. Diaz ◽

...

Keyword(s):

Surgical Management ◽

Performance Metrics ◽

Meniscal Tear ◽

Cohort Analysis ◽

Return To Play ◽

Management Approach ◽

Meniscal Tears ◽

And Performance ◽

Player Performance

Background: Meniscal injuries are extremely common in soccer athletes, and little is known about postrecovery performance. Purpose: To (1) identify characteristics associated with return to play (RTP) to the same league level and (2) evaluate long-term effects that injury and management approach may have on player performance. Study Design: Cohort study; Level of evidence, 3. Methods: Using publicly available records, we identified athletes who sustained meniscal tears across the 5 major European soccer leagues (English Premier League, Bundesliga, La Liga, Ligue 1, and Serie A) between 2006 and 2016. Injured athletes were matched to controls 1:2 by demographics and performance. Investigations included rate of RTP to the same league level, reinjury, player characteristics associated with RTP within 2 seasons, long-term availability, field time, and performance metrics standardized to 90 minutes of play during the next 4 seasons. Results: A total of 250 players sustaining meniscal tears were included, of which 106 (42%) received surgical management. Median absence was 57.5 days (interquartile range [IQR], 35-92) or 7 games (IQR, 4-12). Rate of RTP was 70%, and the reinjury rate 5% if a player could RTP. Age greater than 30 years was a negative predictor for RTP (odds ratio [OR], 0.62; P = .002), whereas higher preinjury goals per game (OR, 2.80; P = .04) and surgical management (OR, 1.38; P = .002) were positive predictors for RTP. Surgical management was associated with higher long-term availability ( P < .01). As compared with the control, there were no significant differences in field time or performance metrics after RTP, either overall or by player position. As compared with nonoperative management, defenders undergoing surgery demonstrated decreased field time. Attackers and midfielders demonstrated similar field time and performance regardless of management. Conclusion: RTP of elite soccer athletes sustaining meniscal tear is contingent on age, preinjury performance, and management approach. Those who RTP to the same league level can be expected to demonstrate equivalent field time, performance, and long-term availability as noninjured athletes.

Download Full-text

Machine Learning for Detecting Data Exfiltration

ACM Computing Surveys ◽

10.1145/3442181 ◽

2021 ◽

Vol 54 (3) ◽

pp. 1-47

Author(s):

Bushra Sabir ◽

Faheem Ullah ◽

M. Ali Babar ◽

Raj Gaire

Keyword(s):

Machine Learning ◽

Performance Metrics ◽

Data Driven ◽

Feature Engineering ◽

Body Of Knowledge ◽

Large Size ◽

Model Training ◽

And Performance ◽

Size Evaluation ◽

And Behavior

Context : Research at the intersection of cybersecurity, Machine Learning (ML), and Software Engineering (SE) has recently taken significant steps in proposing countermeasures for detecting sophisticated data exfiltration attacks. It is important to systematically review and synthesize the ML-based data exfiltration countermeasures for building a body of knowledge on this important topic. Objective : This article aims at systematically reviewing ML-based data exfiltration countermeasures to identify and classify ML approaches, feature engineering techniques, evaluation datasets, and performance metrics used for these countermeasures. This review also aims at identifying gaps in research on ML-based data exfiltration countermeasures. Method : We used Systematic Literature Review (SLR) method to select and review 92 papers. Results : The review has enabled us to: (a) classify the ML approaches used in the countermeasures into data-driven, and behavior-driven approaches; (b) categorize features into six types: behavioral, content-based, statistical, syntactical, spatial, and temporal; (c) classify the evaluation datasets into simulated, synthesized, and real datasets; and (d) identify 11 performance measures used by these studies. Conclusion : We conclude that: (i) The integration of data-driven and behavior-driven approaches should be explored; (ii) There is a need of developing high quality and large size evaluation datasets; (iii) Incremental ML model training should be incorporated in countermeasures; (iv) Resilience to adversarial learning should be considered and explored during the development of countermeasures to avoid poisoning attacks; and (v) The use of automated feature engineering should be encouraged for efficiently detecting data exfiltration attacks.

Download Full-text

Traumatic Leg Fractures in UEFA Football Athletes: A Matched-Cohort Analysis of Return to Play, Reinjury, Player Retention, and Performance Outcomes

Orthopaedic Journal of Sports Medicine ◽

10.1177/23259671211024218 ◽

2021 ◽

Vol 9 (9) ◽

pp. 232596712110242

Author(s):

Ophelie Lavoie-Gagne ◽

Matthew F. Gong ◽

Sumit Patel ◽

Matthew R. Cohn ◽

Avinaash Korrapati ◽

...

Keyword(s):

Performance Metrics ◽

Cohort Analysis ◽

Performance Outcomes ◽

Return To Play ◽

Professional Soccer ◽

Significant Difference ◽

And Performance ◽

Player Performance ◽

Leg Fracture

Background: The average professional soccer team experiences 1 to 2 traumatic leg fractures per season, with unknown effects on player performance. Purpose: To (1) determine the rate and time to return to play (RTP) following leg fracture, (2) investigate the rate of reinjury following RTP, and (3) investigate long-term effects that lower extremity (LE) fracture may have on elite soccer player performance. Study Design: Cohort study; Level of evidence, 3. Methods: Using publicly available records, we identified athletes sustaining a traumatic leg fracture across the 5 major European soccer leagues (English Premier League, Bundesliga, La Liga, Ligue 1, and Serie A) between 2000 and 2016. Athletes with leg fracture (femur, tibia, and/or fibula) were matched 1:2 to controls by demographic characteristics and performance metrics 1 season before the index timepoint. Investigations included the RTP rate, reinjury rate, player characteristics associated with RTP within 2 seasons, long-term player retention, performance metrics during the 4 following seasons, and subgroup analysis by player position. Results: A total of 112 players with LE fracture and 224 controls were identified. Players with LE fractures were absent for a mean of 157 days (range, 24-601 days) and 21 games (range, 2-68 games). The rate of RTP within 1 season was 80%, with 4% experiencing subsequent refracture. Injured players remained active in the league at a higher rate than their uninjured counterparts. As compared with controls, injured athletes played 309 fewer total minutes ( P < .05), scored 0.09 more assists per game ( P < .01) 1 season after injury, and scored 0.12 more points per game 4 seasons after injury ( P < .01). Defenders were most affected by an LE fracture, playing 5.24 fewer games ( P < .05), 603 fewer total minutes ( P < .01), and recording 0.19 more assists per 90 minutes of play as compared with controls 1 season after injury ( P < .001). Attackers and midfielders demonstrated no significant difference in metrics after RTP when compared with controls. Conclusion: Most players sustaining an LE fracture returned to elite soccer at the same level after a significant loss of playing time, with a 4% rate of refracture. Player retention was higher for those sustaining an LE fracture versus uninjured controls. Overall, injured players did not experience a decline in performance after recovery from an LE fracture.

Download Full-text