Session Outline
Anomalies, or outliers, occur in a wide range of applications, and their detection may be significant, for example, for diagnosing diseases, tracking criminal actions, controlling production pipelines, or uncovering operator network changes. In general, the anomalies are contained in unlabelled data, and various unsupervised learning algorithms have been developed for uncovering deviant observations that do not conform to the expectations. However, due to the lack of ground-truth labels, it remains challenging to evaluate how well the anomalies are captured by the model. This presentation discusses the challenges in evaluating whether the machine learning model finds the most significant anomalies focusing on a case study on operator network data.
Key Takeaways
- It is important to characterise a normal observation, to be able to decide what is not normal
- It is generally beneficial to examine the data using various anomaly detection approaches
- Simulation studies can assist in comparing the capabilities of the different methods
————————————————————————————————————————————————————
Speaker Bio
Viivi Uurtio – Data Scientist | Elisa
Viivi Uurtio is a doctor in computer science, and currently works as a data scientist at Elisa. Her expertise entails unsupervised learning in high-dimensional spaces. She develops computational tools for mobile network automation. Until now, she has advanced anomaly detection in operator network data. During her time as a researcher in a machine learning laboratory at Aalto University, she presented her articles in International Conference on Machine Learning 2019 (LA, California, USA), IEEE International Conference on Data Mining 2018 (Singapore), and Discovery Science 2015 (Banff, Alberta, Canada). Additionally, she acted as the first author in a comprehensive tutorial paper on canonical correlation methods published in the highly esteemed journal ACM Computing Surveys.
Day 2 | M8 | Machine and Deep Learning Stage
Viivi Uurtio – Data Scientist | Elisa