On August 13th I gave a talk on „Spatial Data Mining in Precision Agriculture“ at NTNU Trondheim, IDI. Those are the slides. The talk revolved around the two key topics of my to-be PhD thesis: spatial regression and spatial clustering. It generally emphasises the usage of spatial data mining and machine learning methods for spatial data and also introduces to the rather new application field of precision agriculture as a nice side-effect.

I’ve recently been active on the R-help mailing list because I had some issues
with the default implementation of neural networks (nnet). Seems as if the
mailing list solved my problem or at least hinted me towards a solution. The
nnet function seems somewhat strict about its arguments.

http://www.mail-archive.com/r-help@r-project.org/msg118338.html
http://www.mail-archive.com/r-help@r-project.org/msg119643.html

I’ve been nominated to serve as a country representative for Germany in the International Society of Precision Agriculture, which was founded more or less while I was at the ICPA2010 conference in Denver this year. Seems to take off soon.

I’ve been offline for around three months, some kind of mini-sabbatical due to having too many leftover vacation days. That made for a nice work vacation around Trondheim, seeing the hills or mountains around there. The linked tracks are just a few of the recorded ones. Throughout the week I was working at NTNU-IDI, which is more or less the best location to work at I’ve seen so far. The folks around there are friendly, I made contact with some Norwegians and I just loved being there.

I gave an invited talk on August 13th at IDI, which provided me with another chance to talk about my PhD thesis and its two major contributions. Overall, the underlying issue that I’m going to tackle is the specialty of the spatial data I have. It’s partly an experience of „what’s special about spatial data mining?“
Read the rest of this entry »

Four yellow Ortliebs

I’m taking ten weeks off, starting tomorrow. I’ll have my bike with me,
four yellow Ortlieb bags and a small backpack. Direction: North.

I’ll be back in the middle of October, when the winter term starts.

On Tuesday I (and nine other students, all from the U.S.) finally got my ICPA
outstanding graduate student award at the ceremony held during lunchtime. I
haven’t been able to get an appropriate photo, but below there’s the award. I’m
not showing the cheque, though :-) My vision of Precision Agriculture which I had
to describe to receive this award has been posted here.

ICPA outstanding graduate student award

Georg Ruß, Raj Khosla, at ICPA 2010, Denver, Colorado

Raj Khosla, Georg Ruß, Dwayne Westfall, at ICPA 2010, Denver, Colorado

The above photos are from http://www.flickr.com/photos/zimmcomm/sets/72157624407942663/, slightly cropped, scaled and edited. The originals are here: photo1, photo2. There are more photos of the conference in that flickr album.
Read the rest of this entry »

I’m currently at Denver, Colorado, for the 10th International Conference on
Precision Agriculture. So far, it’s been quite interesting to see lots of talks
on what’s (for me) data analysis problems. It’s also nice to see that basic
linear regression is usually the tool which is being used as the most advanced
tool for any kind of prediction tasks. I’ll have too see whether throwing more
advanced data mining stuff at the existing problems is doing any good.

Presenting my talk at the ICPA 2010

My talk on the hierarchical spatial clustering I’ve developed for the purpose
of management zone delineation worked out okay. I think that I did a good job
on adapting my talk to this totally different audience, judging from the
feedback I received after the talk. It was really nice not having to explain
too much details on the data I have because the audience just knew those
attributes. I might even have gotten the point across about what the advantages
of my clustering are in comparison to existing approaches. The presentation slides
are here: russ2010icpa-slides.pdf

Read the rest of this entry »

I’m currently at ICDM in Berlin, the conference which took place in Leipzig in the past two years. Apart from the different location at Alexanderplatz, the quality is the same, and the conference is again very nice. Now that I’m a regular participant, I know a lot of the other people, which is nice if you want to talk to them without having a lot of introduction to do.

My work presented here is a continuation and extension of the IPMU work presented in Dortmund two weeks ago. Again, the emphasis is on getting data mining people into precision agriculture — they’re really needed there. The other aspect of my work is to make sure that spatial data are treated with spatial models, otherwise a lot of the assumptions for non-spatial models are violated and lead to misleading results.

In conjunction with the ICDM I’m holding my workshop on Data Mining in Agriculture for the first time. It’s going to be held this afternoon and so far I have only seen one of the three other presenters. The author of the book Data Mining in Agriculture, Antonio Mucherino, told me that he’s not about to come for personal, urgent reasons, which is a pity, but acceptable.

Some links to the above work: ICDM paper (in Springer LNAI series), DMA workshop paper, the workshop proceedings (of which I’m a co-editor).

Just a quick post that aims to make tomorrow’s slides for my IPMU 2010 talk available. It’s going to be about the management of spatial information and especially the issues which
arise when using non-spatial models on spatial data.

Slides link: russ2010ipmu-slides.pdf

As usual, the paper is in our publication database: Data Mining in Precision Agriculture: Management of Spatial Information.

I’ve had a whole lot of fun writing a paper for the IEEE ICDM conference, which
is going to take place in Sydney, Australia, this year. The programming work
was there, I had some novel data sets to analyse and I came to some cool
conclusions using my homebrew algorithm which explicitly assumes spatial
autocorrelation in the data sets. I could also show that the algorithm produces
meaningless results when spatial autocorrelation does not exist.

It also implements a more or less standard hierarchical agglomerative
clustering procedure on spatial data — there just was no existing work which
fit the problem and the data set, so I had to create my own algorithm using a
straightforward and easily explainable divide-and-conquer approach. I hope that
my reviewers at the IEEE ICDM conference like the idea.

I’m still looking for an easily pronouncable acronym, maybe HACSAD-PA
will do: hierarchical agglomerative clustering for spatially autocorrelated
data from precision agriculture :-)

Read the rest of this entry »