Papers
16
Total Citations
2,089
H-Index
11
About
Eamonn Keogh is a pioneering force in time series data mining, whose work has fundamentally shaped how we analyze and understand sequential data across science and industry. His research centers on developing efficient algorithms for time series similarity, motif discovery, and classification—solving core problems that have enabled breakthroughs in fields from robotics and medicine to climatology and entomology. Keogh’s most celebrated contribution is Derivative Dynamic Time Warping (2001, over 1,100 citations), which dramatically improved the classic DTW algorithm by incorporating derivative information for more robust and faster alignment of time series. He also introduced the concept of time series shapelets (2011, 273 citations), highly discriminative local patterns that revolutionized interpretable classification. More recently, his Matrix Profile series (2016–2020, with papers garnering 161, 77, 62, and 58 citations) has provided a unifying, scalable framework for motif and anomaly detection, now a standard tool in the field. Keogh’s work is characterized by its practical impact—his algorithms are widely deployed in real-world systems, from streaming sensor data classification to texture analysis. A prolific innovator, he has earned numerous best paper awards and is recognized as a leading authority whose ideas continue to inspire a new generation of data mining researchers.
Research Focus
Key Achievements
Top Papers
- 1Derivative Dynamic Time Warping1,124 citations · 2001
- 2Logical-shapelets273 citations · 2011
- 3Detecting time series motifs under uniform scaling168 citations · 2007
- 4Extracting Optimal Performance from Dynamic Time Warping161 citations · 2016
- 5Matrix Profile X77 citations · 2018
- 6
- 7A compression‐based distance measure for texture59 citations · 2010
- 8Matrix Profile V58 citations · 2017
- 9Real-Time Classification of Streaming Sensor Data32 citations · 2008
- 10A Compression Based Distance Measure for Texture25 citations · 2010