There was a town in South America two miles from the river. Every day, the farmers and villagers walked all the way to the river and dragged their loaded buckets all the way back to their homes.
However, there was a sole villager– let’s call him Vladimir– who, instead of dragging a bucket, carried a shovel. And every day, Vladimir dug in the ditches. Through rain, thirst, or hunger: Vladimir just kept digging.
One day, some villagers noticed that Vladimir hadn’t entered town for weeks. The villagers, having had enough of Vladimir’s nonsense, rushed angrily to Vladimir’s house … only to find a small stream flowing right through Vlad’s backyard.
Finding a solution is easy. Every villager found a way to get water. Finding a sustainable, scalable solution: that’s the real challenge. The overarching purpose of technology seems to boil down to two categories: creating resources, and efficiently connecting people to these resources.
Today, we have an abundance (a river?) of resources. I believe in the integration of technology within education, in rapid prevention, and that the Star Wars Prequels are an abomination (maybe this last one isn’t very relevant). The only problem is that the available solutions to these problems, and many others, are restricted by both fiscal and physical obstacles: the river is a lot farther than two miles away.
On this site, I will document my attempts to dig the streams that can connect the amazingly detailed research and solutions to problems in our society to the people who actually face these problems. My current work focuses on the intersection between natural language processing, signal analysis, and computer vision. Specifically, on the task of tying all three complex data representations into concrete emotional effects. I am a Bayesian at heart, and believe that interpretability and true probabilistic modelling always trump slight gains in accuracy (though they often go hand-in-hand).
On this site, I review interesting research I come across, discuss obscure (or often misinterpreted) statistical methods, and post anything else that I just find cool or fun (For more insight regarding my professional work, I’ll redirect you here).
With TensorFlow in one hand, Sci-kit learn in the other, I spend every day digging paths through a flurry data. And I have a smile the whole way through :).
My name is Kian Ghodoussi, and I’m a stream builder. Let’s get started.