Dr Sevvandi Kandanaarachchi
Sevvandi did her PhD in Geometric Analysis, which is a branch of Pure Mathematics at Monash in 2011. After a stint of teaching in Singapore, she got interested in statistical analysis and pursued a Graduate Certificate in Data Mining and Applications from Stanford. While in Singapore, she collaborated with Prof. Kate Smith-Miles and Dr. Shannon Ryan of the DST Group, on a research project which was on the regular bombardment of space debris on space vehicles. After returning to Melbourne, she started working on an industry project under the direction of Prof. Smith-Miles, which has applications related to defence. She is currently working with Prof. Rob Hyndman on anomaly detection and related research areas. As an inter-disciplinary researcher, she also likes to apply her knowledge to real world problems.
Research Interests:
Anomaly Detection
Machine Learning
Meta-Learning
Statistical Learning
Time Series Analysis
Qualifications:
PhD in Mathematics
Graduate Certificate in Data Mining and Applications
Bachelor of Science in Engineering
Projects
Publications
Invited talks, refereed proceedings and other conference outputs
(2021). Here is the anomalow-down!.
useR! 2021.
(2021). Mathematics of anomalies.
Australian Statistical Society and New Zealand Statistical Association Conference.
(2021). Looking out for anomalies.
QUT Mathematics Seminar.
(2021). Lookout! Persisting anomalies ahead..
2021 Australian and New Zealand Statistical Conference (ANZSC2021).
(2021). Evaluating algorithms using Item Response Theory.
RMIT Bundoora Early-Mid Career Researchers (EMCR) Symposium.
(2021). Anomalies and Events keep us on our toes.
Shaping the Future of Data Science: Research Spotlight Series.
(2019). Early event classification in spatio-temporal data streams.
39th International Symposium on Forecasting.
(2019). DOBIN - Dimension Reduction for Outlier Detection.
WOMBAT 2019.
(2019). Instance Space Analysis for Unsupervised Outlier Detection.
1st Workshop on Evaluation and Experimental Design in Data Mining and Machine Learning.
(2018). Outlier Detection in Non-Stationary Data Streams.
Joint International Society for Clinical Biostatistics and Australian Statistical Conference 2018.
(2018). Outlier Detection in Non-Stationary Data Streams.
2018 Joint Statistical Meetings (JSM2018).
(2018). oddstream and stray: Anomaly Detection in Streaming Temporal Data with R.
useR! 2018.
(2018). Anomaly Detection in Non-Stationary Streaming Temporal Data.
38th International Symposium on Forecasting.
(2017). Robust anomaly detection in non-stationary data streams.
WIMSIG Conference 2017.
Journal Articles
(2021). Dimension Reduction for Outlier Detection Using DOBIN.
Journal of Computational and Graphical Statistics. 30(1), 204-219. doi: 10.1080/10618600.2020.1807353
(2021). Early detection of vegetation ignition due to powerline faults.
IEEE Transactions on Power Delivery. 36(3), 1324 - 1334. doi: 10.1109/TPWRD.2020.3006553
(2020). Anomaly Detection in Streaming Nonstationary Temporal Data.
Journal of Computational and Graphical Statistics. 29(1), 13-27. doi: 10.1080/10618600.2019.1617160
(2020). Early classification of spatio-temporal events using partial information.
PLOS ONE. 15(8), e0236331. doi: 10.1371/journal.pone.0236331
(2020). On normalization and algorithm selection for unsupervised outlier detection.
Data Mining and Knowledge Discovery. 12, 389-415. doi: 10.1007/s10618-019-00661-z
(2019). Predicting sediment and nutrient concentrations from high-frequency water-quality data.
PLOS ONE. 14(8), e0215503. doi: 10.1371/journal.pone.0215503
(2019). A framework for automated anomaly detection in high frequency water-quality data from in situ sensors.
Science of The Total Environment. 664, 885 - 898. doi: 10.1016/j.scitotenv.2019.02.085
Publicly available softwares
Technical reports and unrefereed outputs
(2018). On normalization and algorithm selection for unsupervised outlier detection.
(2018). Anomaly Detection in Streaming Nonstationary Temporal Data.
