Current Trends in Learning from Data Streams

Speaker:  João Gama – Porto, Portugal
Topic(s):  Artificial Intelligence, Machine Learning, Computer Vision, Natural language processing


Learning from data streams is a hot topic in machine learning and data mining. In this talk, we present three different problems and discuss streaming techniques to solve them. The first problem is the application of data stream techniques to telecommunications fraud detection. We propose an algorithm for the interconnected by-pass fraud problem. This real-world problem requires processing high-speed telecommunications data and providing fraud alarms in real-time. For the second problem, we present an architecture to explain black-box models for predictive maintenance. The explanations are oriented toward equipment anomalies. For the third problem, we present one of the first algorithms for online hyper-parameter tuning for streaming data. The Self hyper-Parameter Tunning (SPT) algorithm is an optimization algorithm for online hyper-parameter tuning from non-stationary data streams. SPT works as a wrapper over any streaming algorithm and can be used for classification, regression, and recommendation.

About this Lecture

Number of Slides:  53
Duration:  45 minutes
Languages Available:  English, Portuguese
Last Updated: 

Request this Lecture

To request this particular lecture, please complete this online form.

Request a Tour

To request a tour with this speaker, please complete this online form.

All requests will be sent to ACM headquarters for review.