Jesse Read

Multi-label classification is suitable in many domains, for example text categorisation (a document belonging to multiple categories), scene classification (each image may have multiple concepts or ojbects within) as well as video and other media, medical diagnosis, and applications in microbiology. The main challenge is detecting and modelling dependencies between labels, without this modelling being too computationally complex to prevent scaling up to large datasets. Multi-label classification is closely related to ranking, multi-target prediction, and structured output prediction.

Many multi-label applications are found in the context of data streams, where data instances arrive continuously in a theoretically-infinite stream, for example in sensor networks, online social media, news feeds, and large deployments of e-mail. In this context, methods must be able to process large volumes of data quickly and learn and make predictions in real time, as well as detect and adapt to concept drift.

I also have an interest in other topics such as graphical models, neural networks, time series, sequence learning and structured prediction. Several of these are closely related to multi-label classification and data streams.

In Aalto University I am involved in the Traffic Sense - Energy Efficient Traffic with Crowdsensing project: route recognition and prediction, with the overall goal of improving efficiency.

Earlier in the Comonsens project in Spain I worked on formulating and implementing a distributed particle filter on very low-power motes for target tracking.

In the project MultiTree - Multi-scale modelling of tree growth, forest ecosystems, and their environmental control, I am working with forestry scientists to model intra-annual growth of pine trees using machine learning methods.