A ticket to AI & Big Data Day is worth its cost. Data Science only touches grand IT businesses in Lviv and most of the conference participants were true amateurs to this magic in a good sense of the word.
NLP and Deep Learning
The first speaker was Volodymyr Getmanskyi with NLP and NLU in detail. What was interesting is that in this first line of the Data Science technologies able to quickly show business value – they still build models from scratch and using the monsters like IBM Watson does not add much value to solve really specific problems
Talking about some other speakers from Eleks – it was very interesting to hear about practical modeling of a system that recognizes what people wear from Sergey Shelpuk and Olha Romaniuk. They work on building and training a Convolutional Neural Network based on 154 layers of Con-Con-MPool design (actually an adjusted architecture from the winner of the ImageNet challenge). It’s quite fascinating that the main training session lasted for three days using TensorFlow on two Nvidia 1080 GPUs and one CPU.
Data Science in Business and Legal
I have visited a business section of the conference to listen to Mamed Khalilov about probs in AI startups. His main message was “Hey my high-tech friends starting a business in AI, forget about ML (or any other buzzword) in stating goals for you and the team, marketing services or a product to your clients – focus on the business value and use ML as a tool to achieve it. And of course, the knowledge of the business field – the expertise of actually understanding the data goes before the latest deep learning trends and other AI techniques.
Ivan Horodyskyy gave us a good heads-up on the legal issues with AI and Robotics in EU in the near future. I was curious to notice that the definition of the robot was taken from Golem, Frankenstein, Chapek’s robot and Asimov’s Laws. The main question was the liability, as it turns out, the robot cannot be liable and everything falls on the user/ trainer/ owner/ manufacturer/ designer chain. Hopefully, the EU institution clearly states that the object of the regulations must have a physical body which is not a concern for a software product as long as it was not placed on some hardware.
My favorite topic of the conference was from Dmytro Peleshko who is working on the computer vision problems. His ideas and results in the background detection, classification, object tracking and recognition became an inspiration for a few new steps in my works related to object movement tracking. We had a very good discussion of the problem of multiple objects intersection and identification in distributed cameras system during which I shared my solution on the shadow suppression problem.
I pretty much liked the conference, the only thing to work on is the frequency of such events. One year means a lot if to count it from the Moore’s Law standpoint.
Andy Bosyi, CEO/Lead Data Scientist MindCraft
Information Technology & Data Science