Leman Akoglu

About

Leman Akoglu is the Heinz College Dean's Associate Professor of Information Systems at Carnegie Mellon University, holding a tenured position. He directs the Data Analytics Techniques Algorithms (DATA) Lab at Heinz College. His research interests broadly encompass data mining, graph mining, machine learning, and knowledge discovery, with a specific focus on anomalies—identifying and characterizing 'what stands out' in large-scale, time-varying, multi-modal data sources through scalable computational methods. Akoglu holds a Ph.D. in Computer Science from Carnegie Mellon University, obtained in 2012, and a B.S. in Computer Science from Bilkent University, completed in 2007. He also holds courtesy appointments at the Machine Learning Department and the Computer Science Department of the School of Computer Science. His work has led to numerous contributions in anomaly detection, outlier detection, hyperparameter sensitivity analysis, graph neural networks, and the development of foundation models for various applications, including healthcare, finance, and social networks. Akoglu is actively involved in research collaborations, keynote speaking engagements, and organizing workshops and conferences, advancing the field of data science and machine learning.

Research topics

• Computer Science
• Artificial Intelligence
• Computer Security
• Machine Learning
• Data Mining
• Mathematics
• Database
• Data science
• Human–computer interaction
• Theoretical computer science

Selected publications

• Beyond Homophily in Graph Neural Networks: Current Limitations and\n Effective Designs

  arXiv (Cornell University) · 2020 · 265 citations

  We investigate the representation power of graph neural networks in the\nsemi-supervised node classification task under heterophily or low homophily,\ni.e., in networks where connected nodes may have different class labels and\ndissimilar features. Many popular GNNs fail to generalize to this setting, and\nare even outperformed by models that ignore the graph structure (e.g.,\nmultilayer perceptrons). Motivated by this limitation, we identify a set of key\ndesigns -- ego- and neighbor-embedding…

  DOI

• The Need for Unsupervised Outlier Model Selection: A Review and Evaluation of Internal Evaluation Strategies

  ACM SIGKDD Explorations Newsletter · 2023-06-22 · 30 citations

  reviewSenior author

  Given an unsupervised outlier detection task, how should one select i) a detection algorithm, and ii) associated hyperparameter values (jointly called a model)? E ective outlier model selection is essential as di erent algorithms may work well for varying detection tasks, and moreover their performance can be quite sensitive to the values of the hyperparameters (HPs). On the other hand, unsupervised model selection is notoriously difficult, in the absence of hold-out validation data with ground-…

  PublisherDOI

• Deep Anomaly Analytics: Advancing the Frontier of Anomaly Detection

  IEEE Intelligent Systems · 2023-03-01 · 7 citations

  articleOpen access

  Deep anomaly analytics is a rapidly evolving field that leverages the power of deep learning to identify anomalies in various datasets. The use of deep anomaly analytics has increased significantly in recent years due to the growing need to detect anomalies in complex data that traditional methods struggle to handle. Deep anomaly analytics has the potential to transform various industries, including, e.g., healthcare, finance, and cybersecurity, by providing valuable insights and helping to diag…

  PublisherOA PDFDOI

• From Explanation to Action: An End-to-End Human-in-the-loop Framework for Anomaly Reasoning and Management

  arXiv (Cornell University) · 2023-04-06 · 3 citations

  preprintOpen accessSenior author

  Anomalies are often indicators of malfunction or inefficiency in various systems such as manufacturing, healthcare, finance, surveillance, to name a few. While the literature is abundant in effective detection algorithms due to this practical relevance, autonomous anomaly detection is rarely used in real-world scenarios. Especially in high-stakes applications, a human-in-the-loop is often involved in processes beyond detection such as verification and troubleshooting. In this work, we introduce…

  PublisherOA PDFDOI

• Trajectory Anomaly Detection with By-Design Complementary Detectors

  Society for Industrial and Applied Mathematics eBooks · 2025-01-01 · 2 citations

  book-chapterSenior author

  Trajectory anomaly detection is critical across a wide range of applications, from traffic control, and wildlife conservation, to public transportation optimization. However, detecting anomalies in trajectory data is challenging due to the diverse nature of anomalies. In this paper, we propose CETrajAD, an ensemble method for trajectory anomaly detection that integrates complementary detectors, each targeting different aspects of trajectory anomalies. Our approach leverages three types of trajec…

  PublisherDOI

Recent grants

• CAREER: A General Framework for Methodical and Interpretable Anomaly Mining

  NSF · $503k · 2016–2022

• III: Medium: Collaborative Research: Collective Opinion Fraud Detection: Identifying and Integrating Cues from Language, Behavior, and Networks

  NSF · $417k · 2016–2019

• CAREER: A General Framework for Methodical and Interpretable Anomaly Mining

  NSF · $192k · 2015–2016

Frequent coauthors

• Christos Faloutsos

  Carnegie Mellon University

  50 shared

• Lingxiao Zhao

  Dalian Maritime University

  28 shared

• Tina Eliassi‐Rad

  24 shared

• Bart Baesens

  24 shared

• Hanghang Tong

  23 shared

• Véronique Van Vlasselaer

  21 shared

• Monique Snoeck

  21 shared

• Bryan Hooi

  17 shared

Education

• Ph.D., Computer Science

  Carnegie Mellon University

  2004

• M.S., Computer Science

  Carnegie Mellon University

  2000

• B.S., Computer Engineering

  Middle East Technical University

  1996

Awards & honors

• Heinz College Dean's Professor for Feb 2019-2022
• Best Research Paper Award, SIAM SDM 2019
• Best Student Machine Learning Paper Runner-up Award, ECML PK…
• NSF CAREER Award, 2015-2020
• Best Research Paper Runner-up Award, SIAM SDM 2016

Similar researchers at Carnegie Mellon University

• Alessandro Acquistih-66
• Jonathan P. Caulkinsh-59
• David Choih-48