Love the problems, not the solutions. We want to see diverse views in problem solving.
What does research mean at Silo?
Silo.AI is a research and researcher driven organisation that values scientific efforts, be they published papers or internal research endeavours. Silo.AI aims not only to establish itself as a high-quality research organisation, but also and especially to enable individual researchers to establish themselves and develop as top researchers.
We are a customer oriented private AI lab, meaning that our work runs parallel to customer projects, and applies their real life cases whenever possible.
MEET OUR
RESEARCHERS
Scientists & researchers with Silo.AI have diverse backgrounds and cover a wide range of research areas . Meet us.
Sami Remes
AI Scientist
Emil Eirola
AI SCIENTIST
Kaj-Mikael BJÖRK
Head of Research
SAMI REMES
Sami Remes represents Silo.AI’s new generation of researchers. He began his career in AI at the Aalto University in Helsinki. Already during his BSc degree, he began working in Professor Samuel Kaski’s Statistical Machine Learning and Bioinformatics Group , which is nowadays part of Kaski’s Probabilistic Machine Learning Group.
Sami completed both his BSc and MSc theses on extensions of the Group Factor Analysis model, developed by his research group, publishing his findings in the workshops of the world’s leading AI conference NIPS in 2013 and 2015. He then moved on to research Gaussian Processes and developing new kernels that model non-stationary changes in e.g. dependencies of different output variables. This research led to various publications in forums such as ACML (Asian Conference on Machine Learning) and the main conference of NIPS in 2017.
Upon joining Silo.AI, Sami has finished most of his PhD research and will receive his degree in early 2019.
Outside the office, Sami enjoys a married life, barbecue, and the occasional game on PC or PS4.
Learn about Sami’s research on non-stationary spectral kernels.
EMIL EIROLA
Emil Eirola is one of the first expert hires in Silo.AI, bringing machine learning experience to the company. With a Master’s thesis (2009) on feature selection and PhD thesis (2014) on Machine Learning for data with missing values, his research has focused on how to deal with the practical issues of applying ML to real-world problems. An undergraduate major in mathematics ensures a strong understanding of the theoretical foundations behind the algorithms.
As a researcher in the Applications of Machine Learning group in Aalto University, and later the Department of Business Management and Analytics at the Arcada University of Applied Sciences, Emil has worked on finding machine learning solutions to all sorts of different use cases in fields such as finance, security, and healthcare.
When not in the office, Emil can be found travelling the world with any one of his several music projects.
KAJ-MIKAEL BJÖRK
Kaj-Mikael Björk is a top senior researcher with unique capability to lead research teams. Kaj-Mikael is Research Director at Arcada and Head of Research of Silo.AI. During his time as the Head of Department in Arcada, he initiated the AI track and Risklab Finland. Within the research projects he has participated in, he has completed more than 70 scientific peer reviewed articles with an H-index of 13 (Google scholar). He was also listed as the top 6 most published AI researcher in Finland in the Digibarometri survey by the Research Institute of the Finnish Economy ETLA.
Previously, Kaj-Mikael has been a visiting professor at UC Berkely and Carnegie Mellon University as well as working as an Assistant Professor in Information Systems (Åbo Akademi University) and Senior Lecturer in Logistics (Arcada). The borderline between economics and IT has long fascinated him, as well as inspired him to pursue achievements in both education and research. He has held approx. 15 different courses in the fields of Machine Learning, Logistics, Management Science and Engineering. As Head of Department, Kaj-Mikael has also participated actively in many administrative tasks, such as a member of the university’s steering group and other task forces. His research interests are in optimization, machine learning, analytics, supply chain management and fuzzy logic.
Learn more about the Silo.AI Journal Club that Kaj-Mikael is hosting.
SNAPSHOT OF our
researchers' PUBLICATIONs
SILO.AI SCIENTISTS ARE PRODUCTIVE, WELL-PUBLISHED SCHOLARS
State Space Gaussian Processes with Non-Gaussian Likelihood
Nickisch, H., Solin, A., Grigorevskiy, A., 2018. State Space Gaussian Processes with Non-Gaussian Likelihood. ICML 2018.
Learning structures of Bayesian networks for variable groups
Parviainen, P., Kaski, S., 2017. Learning structures of Bayesian networks for variable groups. IJAR 88.
Non-Stationary Spectral Kernels
Remes, S., Heinonen, M., Kaski, S., 2017. Non-Stationary Spectral Kernels. NIPS 2017.
Bank distress in the news: Describing events through deep learning
Rönnqvist, S., Sarlin, P., 2017. Bank distress in the news: Describing events through deep learning. Neurocomputing 264.
Advanced Query Strategies for Active Learning with Extreme Learning Machine
Akusok, A., Eirola, E., Miche, Y., Gritsenko, A., Lendasse, A., 2017. Advanced Query Strategies for Active Learning with Extreme Learning Machine. ESANN 2017.
Adding reliability to ELM forecasts by confidence intervals
Akusok, A., Gritsenko, A., Miche, Y., Björk, K-M., Nia, R., Lauren, P., Lendasse, A., 2017. Adding reliability to ELM forecasts by confidence intervals. Neurocomputing 219.
Toward robust early-warning models: A horse race, ensembles and model uncertainty
Holopainen, M., Sarlin, P., 2017. Toward robust early-warning models: A horse race, ensembles and model uncertainty. Quantitative Finance 17.
Structure discovery in Bayesian networks by sampling partial orders
Niinimäki, T., Parviainen, P., Koivisto, M., 2016. Structure discovery in Bayesian networks by sampling partial orders. JMLR 17.
Fast zeta transforms for lattices with few irreducibles
Björklund, A., Husfeldt, T., Kaski, P., Koivisto, M., Nederlof, J., Parviainen, P., 2016. Fast zeta transforms for lattices with few irreducibles. TALG 2016.
Global optimisation of a portfolio adjustment problem under credibility measures
Lundell, A., Björk, K-M., 2016. Global optimisation of a portfolio adjustment problem under credibility measures. IJOR 25.
High-Performance Extreme Learning Machines: A Complete Toolbox for Big Data Applications
Akusok, A., Björk, K-M., Miche, Y., Lendasse, A., 2015. High-Performance Extreme Learning Machines: A Complete Toolbox for Big Data Applications. IEEE Access 3.
Tractable Bayesian Network Structure Learning with Bounded Vertex Cover Number
Korhonen, J., Parviainen, P., 2015. Tractable Bayesian Network Structure Learning with Bounded Vertex Cover Number. NIPS 2015.
SILO.AI RESEARCH NETWORK
is a 350+ strong community of Machine Learning researchers around the world.
Snapshot of the Silo.AI Research Network
TOMASTO ASTE
professor
Univ college london
Jukka Corander
Professor
university
of oslo
Moncef Gabbouj
PROFESSOR
Tampere univ of tech
Guang bin Huang
PRofessor
nanyang tech
university
Samuel Kaski
PROFESSOR
Aalto University
Matti Pietikäinen
PROFESSOR
University of Oulu
Scott Sanner
PROFESSOR
University of toronto
Interested in joining the Silo.AI Research Network on Slack?