Tapio Pahikkala
Professor
aatapa@utu.fi +358 29 450 4323 +358 50 345 5824 Agora Office: 456D ORCID identifier: https://orcid.org/0000-0003-4183-2455 |
Machine learning, Data science, Artificial intelligence
Tapio Pahikkala is a professor of computer science in the University of Turku, Finland, from which he also received his doctoral degree in 2008. He has authored more than 150 peer-reviewed scientific articles and participated in the winning teams of several international scientific competitions/challenges. He has led many research projects, supervised more than ten doctoral theses, held several positions of trust in academia and served in the program committees of numerous international conferences. His current research interests include theory and algorithmics of machine learning, data analysis, and artificial intelligence, as well as their applications on various different fields.
Theory and algorithmics of machine learning, data science and artificial intelligence as well as their practical applications in various different fields. Estimation of prediction performance with resampling methods, theory of resampling and cross-validation.
Current research projects:Academy of Finland: "AI technologies for interaction prediction in biomedicine", Academy of Finland: "Machine Learning for Systems Pharmacology", Business Finland: "Privasa".
The course I am currently responsible of: ``Evaluation of Machine Learning Methods'', consists of a series of practical cases studies that are each presented by different assistant teachers that act as clients of data scientists. The clients then introduce the problem the the data scientist should solve for them and the details of the data. The students' job is then implement the data analysis pipeline, train a predictive model, do a proper experimental design and carry out carry out statistical estimation of the prediction performance for each client. To achieve this, they study the accompanying course material that is currently in the form of both video lectures and reading material. All the clients' cases correspond to real cases from which our team has written research articles in the past. For example, the case concerning metal ion concentration prediction from drinking water is based on our research cooperation with the chemistry deparment of the University of Turku (Pihlasalo et al. 2016), the case on water permeability prediction in forestry for route planning of forest harvesters and the use of newly developed spatial cross-validation for estimating the prediction performance in that context is based on our cooperation with the Natural Resources Center of Finland (Pohjankukka et al. 2017), and the case concerning drug-target interaction prediction is based on our research cooperation with Institute for Molecular Medicine Finland (Pahikkala et al. 2015), to highlight a few. We have also had plans to involve cases from private companies in the future, such that would correspond to real commercial cases.
- A comparison of embedding aggregation strategies in drug-target interaction prediction (2024)
- BMC Bioinformatics
(Refereed journal article or data article (A1)) - Predicting pairwise interaction affinities with ℓ0-penalized least squares-a nonsmooth bi-objective optimization based approach∗ (2024)
- Optimization Methods and Software
(Refereed journal article or data article (A1)) - Budget-based classification of Parkinson's disease from resting state EEG (2023)
- IEEE Journal of Biomedical and Health Informatics
(Refereed journal article or data article (A1)) - Empirical evaluation of amplifying privacy by subsampling for GANs to create differentially private synthetic tabular data (2023)
- CEUR Workshop Proceedings
(Refereed article in conference proceedings (A4)) - Evaluating Classifiers Trained on Differentially Private Synthetic Health Data (2023)
- Proceedings (IEEE International Symposium on Computer-Based Medical Systems)
(Refereed article in conference proceedings (A4)) - Bayesian Inference for Predicting the Monetization Percentage in Free-to-Play Games (2022)
- IEEE Transactions on Games
(Refereed journal article or data article (A1)) - Detection of Prostate Cancer Using Biparametric Prostate MRI, Radiomics, and Kallikreins: A Retrospective Multicenter Study of Men With a Clinical Suspicion of Prostate Cancer (2022)
- Journal of Magnetic Resonance Imaging
(Refereed journal article or data article (A1)) - Generalized vec trick for fast learning of pairwise kernel models (2022)
- Machine Learning
(Refereed journal article or data article (A1)) - Quicksort leave-pair-out cross-validation for ROC curve analysis (2022)
- Computational Statistics
(Refereed journal article or data article (A1)) - Adaptive risk prediction system with incremental and transfer learning (2021)
- Computers in Biology and Medicine
(Refereed journal article or data article (A1)) - A Resource Management Model for Distributed Multi-Task Applications in Fog Computing Networks (2021)
- IEEE Access
(Refereed journal article or data article (A1)) - Effect of oat β-glucan of different molecular weights on fecal bile acids, urine metabolites and pressure in the digestive tract – A human cross over trial (2021)
- Food Chemistry
(Refereed journal article or data article (A1)) - Modeling drug combination effects via latent tensor reconstruction (2021) Wang Tianduanyi, Szedmak Sandor,Wang Haishan, Aittokallio Tero, Pahikkala Tapio, Cichonska Anna,Rousu Juho
(Refereed journal article or data article (A1)) - Negative Predictive Value of Biparametric Prostate Magnetic Resonance Imaging in Excluding Significant Prostate Cancer: A Pooled Data Analysis Based on Clinical Data from Four Prospective, Registered Studies (2021)
- European Urology Focus
(Refereed journal article or data article (A1)) - A general-purpose toolbox for efficient Kronecker-based learning (2020) JuliaCon Proceedings Michiel Stock, Tapio Pahikkala, Antti Airola, Bernard De Baets
(Other (O2)) - Algebraic shortcuts for leave-one-out cross-validation in supervised network inference (2020)
- Briefings in Bioinformatics
(Refereed journal article or data article (A1)) - Cost-effective survival prediction for patients with advanced prostate cancer using clinical trial and real-world hospital registry datasets (2020)
- International Journal of Medical Informatics
(Refereed journal article or data article (A1)) - Developing a pain intensity prediction model using facial expression: A feasibility study with electromyography (2020)
- PLoS ONE
(Refereed journal article or data article (A1)) - GeFeS: A generalized wrapper feature selection approach for optimizing classification performance (2020)
- Computers in Biology and Medicine
(Refereed journal article or data article (A1)) - Leveraging multi-way interactions for systematic prediction of pre-clinical drug combination effects (2020)
- Nature Communications
(Refereed journal article or data article (A1))