I am a researcher with passion towards Big Data in ML, hardware accelerated algorithms on local and mobile devices, and especially towards making things that work over talking about things that don't exist. I enjoy sharing my knowledge with others via teaching and supervision.

My background includes BSc in IT at Moscow (2011), MSc in ML and Data at Aalto (2014) and DSc in ML at the University of Iowa, USA (2016). I did postdoc research and created several courses from scratch for the Big Data Analytics program at Arcada. I worked at ML consultancy and as a software/data engineer at a large enterprise.

Research wise, I work a lot with the method called Extreme Learning Machines, or ELM in short. I have 1500 citations and an h-index of 15 as of 2025, find more on Google Scholar (https://scholar.google.com/citation… External link).

Some of the interesting projects I created are "HPELM" library (https://hpelm.readthedocs.io External link) with a few hundred thousand downloads for high-performance ML with GPU acceleration, its continuation "Scikit-ELM" compatible with a popular ML framework Scikit Learn. I wrote an out-of-memory GPU accelerated version of the same ML method that runs on iOS devices like iPhone or iPad. Using this method, I created an app for autonomous on-device learning from maps, and a real time prediction and display of values over maps with application to acid sulfate soils in south-western Finland (https://github.com/akusok/HaSuRiski External link).

For personal interest, I learn the practical applied part of local LLMs, and I like playing creative games like Factorio. I also like travelling around the world with the family.

Find my by arcada email, on Arcada Teams chat, or a personal email E-mail: akusok.a@gmail.com