I was
always fascinated by the vastness of the Space, depth of the oceans and the
darkness of the night. Things we don’t know, yet. And also the knowledge we
already possess thanks to the advance in the science throughout the history of
humanity.
The “star -gazing” has inspired my exploratory data analysis in a way
that I go beyond usual statistical
calculations, further to investigate the hidden dimensions in the data.
I take a use of the Graph theory and Bayes theorem, so the results of my analysis can be used in feature extractions and expert learning for the Machine learning process, by the actual machine learning experts.
The visual outcome holds the high density of the information about the subject, dataset or the topic.
Inspired by the modern Neuroscience, I form the nebula’s to use the natural brain and sight features like the anomaly recognition, which works much better in graphical form than with the numbers.
The process starts with choosing the subject of the study. Life sciences, cybersecurity, social studies are just a few examples of great sources of the ideas for the actual datasets.
Next I generate the dataset – it has to be semantically correct. Imagine the dots, connected with the links.
Let the dot call the node and the link connecting the dots let will be the edge.
Nodes are the nouns and edges are the verbs. This concept helps to tell a story with the picture as well as provide cybernetic concept for the particular data analysis itself.
Which furthermore helps with the automation and may also feed a generation of models for the machine learning.
Every picture is an output of processing the dataset into a form of the connected graph(s) and an application of certain algorithms like Force directed graph drawing.
The whole point of my analysis is to understand better to the studied topics and eventually help the machines/ computers to understand it, too.
In order to provide you with the visual form, I use open source tools like Gephi and NodeXL. I like the Graphistry, too. To do the heavy lifting behind the scene, I work with both community and Enterprise versions of Neo4j database, also with python based libraries in anaconda pack.
To obtain the data I feed to mentioned tools, I use software like Splunk or I code custom crawler / scraper to get the online data for my process.
The
shapes:
As the general laws of the physics are applicable anywhere in our universe, so
they are valid for my visuals. Therefore some of them might remind you of the
stars, planets, various biological structures or even the sea creatures.
Some of my fans call them nebulas, some others perceive it as a flowers, you name it.
It is a consequence of a known phenomenon of the way the visual recognition works in our brains. You tend to assign shapes you see and imagine them to be the objects you tend to like, dislike or even fear. This all is based on the current mood, memories and experience of each and every individual at the moment.
Therefore it is just natural, if you have completely different feeling than your friend, while you enjoy my visual art.
And I really love to observe, how many various impressions is created by my pictures.
——————————
Photo credit:
Michala Rusaňuková –
@Misha.photo
https://www.misha-photography.com/