Protovis – a graphical toolkit for visualization

May 23, 2010

Protovis is a visualization toolkit for JavaScript that takes a graphical approach to data visualization by allowing the user to specify how the data should be encoded in the marks representing it on the screen. Protovis uses SVG image for rendering, allowing visualizations to be seamlessly inserted into pages.

The developers describe the advantages as:

“It retains the conceptual simplicity and low-level control of graphical systems by dealing directly with graphical elements (shapes, lines,i.e., marks), but specifies marks declaratively as encodings of data. Additionally, we provide useful abstractions as optional tools to simplify visualization.”

Here are some examples:




Nightingale´s Rose

Einstein once said

May 23, 2010

“Not everything that can be counted counts, and not everything that counts can be counted.”

To read more about this click on the article Metric Mania by John Allen Paulos for The New York Times.

Who would not agree???

May 22, 2010

Ever heard of Weka?

May 22, 2010

It means Waikato Environment for Knowledge Analysis as it was developed at the University of Waikato (New Zealand), although Weka as well is a bird that is endemic to New Zealand. I like that maybe because I spend some time in New Zealand myself.
However, the Weka software is easy to use due to the graphical user interfaces it contains. It supports several data mining techniques such as preprocessing, classification, clustering, regression and many more. You can download Weka for free on this page.
Sure you will need some tutorials on how to use Weka! Visit this web page for free video tutorials. You will find videos on how to apply different data mining tools like Text Mining, Neural Network, Clustering etc. and you can download the data sets from the tutorials to experiment with them on your own. The quality of the videos is good and the lectures are given in a understandable pace so you should not have problems to follow them. Although the lecturer gets confused sometimes which can be disturbing for the listener. But my favorite tutorials about Weka are from IBM. They show how to use Weka and how to interpret the results as well, which is an advantage if you want to improve your data mining skills. So I recomend to check out the page.

RapidMiner- Data Mining Software

May 22, 2010

If you are looking for a really good data mining software you should consider RapidMiner. RapidMiner is a leading open-source system for knowledge discovery and data mining with a graphical user interface. It supports a variety of data mining algorithms as decision trees, self-organization maps, clustering, classification to name only a few. Of course it can be downloaded for free on the RapidMiner website where you can also find free video tutorials on how to use the software. As it is a wide-spread software you will be able to find many sources and discussion boards about RapidMiner.
A blog with good content about RapidMiner (and more interesting topics) is Neural Market Trends. Check out the video tutorials here!

Free video tutorials from StatSoft

May 22, 2010

One of my favorite free courses of data mining is from StatSoft. You can find the tutorials on youtube. If you want to practice with StatSoft you can download a free 30 day trial version of there Statistica software on the StatSoft page.

There are 35 Sessions which give practical insight on how to apply data mining. I think that this a really good video series if you are at a beginners level because those very professionaly made videos focus on how data mining works in general and not only about how to use there software. After watching those 35 sessions you really get an idea how to do a whole data mining analysis from data cleaning to interpreting the results. The first two sessions give an overview about Data Mining and CRISP. The following sessions are about preparing the data and applying different DM techniques with the help of credit risk data and marketing data respectivly. So if you are a practical learner (and who is not) you should give it a try and just listen to the videos which are about 10 minutes each.

Here you can finde the topics of the videos.