I am a computer scientist, educator, and a former start-up CPO. (Speechly got acquired by Roblox in September 2023.)
Currently I am on sabbatical. Occasionally I teach and do a bit of research or consulting. I am also a Docent (of data science) at the Department of Computer Science, University of Helsinki.
"Kuinka meitä huijataan netissä tekoälyn avulla?" (in Finnish), 4.12.2019 at Tiedekulma in Helsinki.
What is AI and will it take our jobs away?, 17.11.2016 at Oxford Internet Institute in Oxford, England.
Here you can find reference implementations of some algorithms I've worked on, and datasets used in some of my papers.
Triplet data that we collected for the experiments in (Heikinheimo & Ukkonen, HCOMP 2013) is available here. Please see the README file for instructions. If you use this in your research, I am kindly asking you to cite our HCOMP 2013 paper.
Spectra is an algorithm for quickly estimating the "pattern frequency spectrum" of a binary dataset. This is a curve that shows the number of frequent itemsets for a given support threshold.
Please try out a demo, it will work even on your (smart)phone! Source code is available on GitHub.
The paper (van Leeuwen & Ukkonen, ECMLPKDD 2014) is available online, and more details about the demo in particular can be found in our short technical report.
Fast-Skyline is an algorithm for computing approximate “skylines” (non-dominated sets) of subsets of size-k subject to two functions, one linear, one submodular. That is, the algorithm computes the set of non-dominated subsets of size-k. This problem has applications in e.g. viral marketing.
The algorithm is described in (van Leeuwen & Ukkonen, SDM 2015), and extends the work in (van Leeuwen & Ukkonen, ECMLPKDD 2013). More information can be found here.
In our ECMLPKDD 2017 paper we describe a method for ranking subgroups using network-based techniques. The code is available on GitHub.
Cost calculator is a simple Javascript application for learning how asymmetric misclassification costs and imbalanced class distribution affect performance. I made this for teaching purposes, maybe it's useful for you too.
antti.ukkonen (at) helsinki.fi (for teaching and research related things)
antti.ukkonen (at) gmail.com (for other business)
My LinkedIn page