Leman Akoglu

What is she best known for?

The fields of study she is best known for:

• Artificial intelligence
• Machine learning
• Statistics

The scientist’s investigation covers issues in Graph, Data mining, Anomaly detection, Artificial intelligence and Theoretical computer science. Her work carried out in the field of Graph brings together such families of science as Transfer of learning, Community structure and Cluster analysis. Her Data mining study incorporates themes from Graph, The Internet, Spamming, Subnetwork and Information retrieval.

Her work in Anomaly detection addresses issues such as Outlier, which are connected to fields such as Graph size, Change detection and Suite. The Artificial intelligence study combines topics in areas such as Machine learning and Relational database. Her study in Theoretical computer science is interdisciplinary in nature, drawing from both Modular decomposition, Graph theory, Null model and Weighted network.

Her most cited work include:

• Graph based anomaly detection and description: a survey (625 citations)
• OddBall: spotting anomalies in weighted graphs (390 citations)
• RolX: structural role extraction & mining in large graphs (283 citations)

What are the main themes of her work throughout her whole career to date?

Leman Akoglu mainly investigates Anomaly detection, Graph, Artificial intelligence, Data mining and Theoretical computer science. Her Anomaly detection research is multidisciplinary, incorporating elements of Hyperparameter, Unsupervised learning, Outlier and Benchmark. Her research investigates the link between Graph and topics such as Graph that cross with problems in Homophily.

The various areas that Leman Akoglu examines in her Artificial intelligence study include Natural language processing, Machine learning and Pattern recognition. In her study, which falls under the umbrella issue of Data mining, Detector is strongly linked to Ground truth. Her work investigates the relationship between Theoretical computer science and topics such as Modular decomposition that intersect with problems in Indifference graph and Combinatorics.

She most often published in these fields:

• Anomaly detection (32.61%)
• Graph (29.71%)
• Artificial intelligence (27.54%)

What were the highlights of her more recent work (between 2018-2021)?

• Anomaly detection (32.61%)
• Artificial intelligence (27.54%)
• Outlier (14.49%)

In recent papers she was focusing on the following fields of study:

Leman Akoglu mostly deals with Anomaly detection, Artificial intelligence, Outlier, Graph and Machine learning. Her study on Anomaly detection also encompasses disciplines like

• Benchmark together with Data mining, Curse of dimensionality and Inference,
• Hyperparameter together with Model selection. Her study looks at the intersection of Data mining and topics like Distributed Computing Environment with Key.

Her work on Cluster analysis, Class and Classifier is typically connected to Scale as part of general Artificial intelligence study, connecting several disciplines of science. Her biological study spans a wide range of topics, including Graph, Theoretical computer science and Pattern recognition. Her study in the fields of Interpretability under the domain of Machine learning overlaps with other disciplines such as Model building.

Between 2018 and 2021, her most popular works were:

• PairNorm: Tackling Oversmoothing in GNNs (55 citations)
• PairNorm: Tackling Oversmoothing in GNNs (15 citations)
• Statistical Analysis of Nearest Neighbor Methods for Anomaly Detection (10 citations)

In her most recent research, the most cited papers focused on:

• Artificial intelligence
• Machine learning
• Statistics

Her primary areas of study are Graph, Graph neural networks, Algorithm, Artificial neural network and Normalization. Her Graph study combines topics in areas such as Theoretical computer science, Convolutional neural network, Heterophily, Homophily and Perceptron. Her research on Graph neural networks concerns the broader Graph.

Her Algorithm study frequently draws parallels with other fields, such as Network architecture.