Machine Learning – A Q&A With our our very own Data Scientist
Machine Learning – A Q&A With our very own Data Scientist – Sergii Nechuiviter
Name: Sergii Nechuiviter
Title: Senior Data Science Engineer
Favorite Food: fried potatoes
Favorite Music Genre: epic music from movies and games
Where and what did you study at university?
Moscow Institute of Physics and Technology (State University)
Wide study of physics and deep study of machine learning.
Who are the influences in your professional life and why?
I learned a lot from my MIPT lecturer of machine learning – Konstantin Vorontsov. He gave me a deep understanding of theoretical and practical principles of data mining. I returned to those lectures and seminars multiple times during my career.
What got you interested in pursuing data science as a career?
Actually, I am interested in Computational Intelligence. But, being a more practical than theoretical person, I like to work in the field of data science and machine learning as it involves me more tightly in the real implementation of machine intelligence.
What are some of the projects you currently working on at display.io?
At display.io I am involved in the engineering of new predictors of the probability of conversion for the given ads.
What are the techniques you are using to solve these challenges?
For training of those predictors the Python-based tools are used: numpy for linear algebra, scikit-learn for some simple machine learning, dask for pure python distributed computations and resource management and our own implementation of more complex machine learning algorithms like FFM.
In your opinion what are the issues affecting mobile marketing that applied data science can improve?
I think applied data science boosts and speeds up abilities of “classic” analytics to select the best marketing strategy. With machine learning tools data scientist can analyze available data deeper: makes audience targeting more precise, quicker reacts to changes in trends and moods.