I’ll be giving a talk at the Australian Taxation Office, likely to take place on 11th of December at 1100 local time (Canberra, ACT). The slides can be obtained here: slides-russ2009ato.pdf. The abstract is as follows:

Data Mining in Agriculture

In recent years, due to new and affordable technological advances, data
collection has turned into an everyday task. Nowadays, especially with the
advent of the global positioning system and modern farming vehicles, sensors
and equipment, even agriculture has turned into a data-driven discipline
called precision agriculture. However, as in numerous other research and
production areas, collecting data is not sufficient for economic or ecological
well-being. The collected data have to be mined and turned into usable
knowledge.

Therefore, this talk presents some approaches towards data mining in
agriculture. The talk will begin with a short overview about the origins of
the actual agriculture data. The difference between spatial and non-spatial
approaches will be emphasised using an example of yield prediction. Some of
the non-spatial techniques such as clustering, regression and feature
selection may be carried over to spatial approaches. Most of the presented
work considers very recent issues which remain unsolved in this discipline so
far (at least to the speaker’s knowledge). Furthermore, the presented work is
an excerpt of what is going to be the speaker’s PhD thesis, which is likely to
cover Data Mining in Agriculture from a computer scientist’s perspective.