Pekka Marttinen

Postal Address:

Helsinki Institute for Information Technology HIIT, Department of Computer Science

Aalto University

P.O.Box 15400

FIN-00076 AALTO, FINLAND

Street Address:

Room A317, Konemiehentie 2, Espoo, Finland

Email:

pekka.marttinen (at) aalto.fi

---

Research Group: Machine Learning for Health (Aalto-ML4H)

• Researchers

  • Pekka Marttinen, PI
  • Hans Moen, Research Fellow
  • Ti John, Research Fellow
  • Minttu Alakuijala, Postdoc (jointly with the PML group)
  • Alexander Nikitin, Postdoc
  • Negar Safinianaini, Postdoc
  • Linli Zhang, Postdoc
  • Zhiyuan Li, Postdoc (jointly with Joni Pajarinen)

• PhD students

  • Caglar Hizli
  • Antti Pöllänen
  • Joel Honkamaa
  • Fanni Ojala
  • Tommi Gröhn
  • Ya Gao
  • Matias Paloranta (with Antti Rannikko at U Helsinki)
  • Co-supervisor with Alexander Ilin in the Deep Learning group:
    • Nicola Dainese
    • Kalle Kujanpää
    • Sam Spilsbury
    • Arturs Polis
  • Co-supervisor with Rohit Babbar in the Extreme Classification group:
    • Erik Schultheis
    • Jinbin Zhang
    • Nasibullah Nasibullah

• Research assistants

  • Mihriban Kocak

• Graduated PhD students

  • Tomi Peltola (2014). Sparse Bayesian Linear Models - Computational Advances and Applications in Epidemiology. Supervisors: Aki Vehtari, Pekka Marttinen
  • Marko Järvenpää (2020). Gaussian Process Surrogate Methods for Sample Efficient Approximate Bayesian Computation. Supervisors: Pekka Marttinen, Aki Vehtari
  • Tianyu Cui (2023). Interaction Detection with Probabilistic Deep Learning for Genetics. Supervisors: Samuel Kaski, Pekka Marttinen
  • Shaoxiong Ji (2023). Natural Language Processing for Healthcare: Text Representation, Multitask Learning, and Applications. Supervisor: Pekka Marttinen
  • Yogesh Kumar (2024). Efficient Transfer Learning with Sequential and Multi-Modal Approaches for Electronic Health Records. Supervisor: Pekka Marttinen
  • Mohammadreza Mohammadnia Qaraei (2024). Efficient and robust algorithms for extreme multilabel classification. Supervisors: Rohit Babbar, Pekka Marttinen
  • Katsiaryna Haitsiukevich (2025). Advances in physics-informed deep learning. Supervisors: Alexander Ilin, Pekka Marttinen
  • Onur Poyraz (2025). Flexible Bayesian latent variable modeling of interacting processes in healthcare time-series. Supervisor: Pekka Marttinen

• Former Postdocs

  • Reza Ashrafi, Santosh Hiremath, Vishnu Raj, Prayag Tiwari

• Former group members (Master's students, visitors, etc)

  • Guangyi Zhang, Maximilian Proll, Erick Gijon, Alejandro Ponce de Leon Chavez, Berrenur Saylam, Severi Rissanen, Juuso Kylmäoja, Teijo Väärä, Nicola Brazzale, Priscilla Ong, Jasu Vehtari, Shinji Hotta, Arina Odnoblyudova, Anlin Sun, Matteo Merler, Linh Nguyen, Jouni Tammenmaa

Research Interests

• Machine learning methodology

  • Large Language Models
  • Reinforcement learning
  • Deep learning
  • Probabilistic Machine Learning
  • Causal inference

• Applications

  • Healthcare
  • Bioinformatics
  • Statistical genetics
  • Epidemiology
  • Personalized medicine

Publications

Full list of publications

Google Scholar

Selected Articles (Probabilistic Machine Learning, Causality)

• Hizli, C., John, S.T., Juuti, A., Saarinen, T., Pietiläinen, K., and Marttinen, P. (2023). Temporal causal mediation through a point process: direct and indirect effects of healthcare interventions. Thirty-seventh Conference on Neural Information Processing Systems (NeurIPS 2023). Available online

• Hizli, C., John, S.T., Juuti, A., Saarinen, T., Pietiläinen, K., and Marttinen, P. (2023). Causal Modeling of Policy Interventions From Treatment–Outcome Sequences. In Proceedings of the 40th International Conference on Machine Learning, PMLR 202. (ICML 2023). Available online

• Raj, V., Cui, T., Heinonen, M., and Marttinen, P. (2023). Incorporating functional summary information in Bayesian neural networks using a Dirichlet process likelihood approach. In Proceedings of 26th International Conference on Artificial Intelligence and Statistics, PMLR 206:6741-6763. (AISTATS 2023) Available online

• Cui, T., Kumar, Y., Marttinen, P., and Kaski, S. (2022). Deconfounded Representation Similarity for Comparison of Neural Networks. Thirty-sixth Conference on Neural Information Processing Systems (NeurIPS 2022). Available online

• Cui, T., Havulinna, A., Marttinen, P.*, and Kaski, S.* (2022). Informative Bayesian Neural Network Priors for Weak Signals. Bayesian Analysis, 17(4):1121-1511. (*equal contribution) Available online

• Rissanen, S. and Marttinen, P. (2021). A Critical Look at the Consistency of Causal Estimation with Deep Latent Variable Models. Thirty-fifth Conference on Neural Information Processing Systems (NeurIPS 2021), pp. 4207-4217. Available online

• Järvenpää, M., Gutmann, M.U., Pleska, A., Vehtari, A., and Marttinen, P. (2019). Efficient acquisition rules for model-based approximate Bayesian computation. Bayesian Analysis, 14(2):595-622. Available online

Selected Articles (Large Language Models)

• Nikitin, A., Kossen, J., Gal, Y., and Marttinen, P. (2024). Kernel Language Entropy: Fine-grained Uncertainty Quantification for LLMs from Semantic Similarities. In Thirty-Eighth Annual Conference on Neural Information Processing Systems (NeurIPS 2024), accepted. Preprint

• Dainese, N.*, Merler, M.*, Alakuijala, M., and Marttinen, P. (2024). Generating Code World Models with Large Language Models Guided by Monte Carlo Tree Search. In Thirty-Eighth Annual Conference on Neural Information Processing Systems (NeurIPS 2024), accepted. (*equal contribution) Preprint

• Spilsbury, S., Marttinen, P., and Ilin, A. (2024). Generating Demonstrations for In-Context Compositional Generalization in Grounded Language Learning. In Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing (EMNLP 2024), accepted. Preprint

• Dainese, N., Marttinen, P., and Ilin, A. (2023). Reader: Model-based language-instructed reinforcement learning. The 2023 Conference on Empirical Methods in Natural Language Processing (EMNLP 2023). Available online

Selected Articles (Computational Biology and Health)

• Gao, Y., Moen, H., Koivusalo, S., Koskinen, M., and Marttinen, P. (2024). Query-Guided Self-Supervised Summarization of Nursing Notes. Proceedings of the Machine Learning for Health Symposium (ML4H 2024), accepted. Preprint

• Honkamaa, J., Khan, U., Koivukoski, S., Valkonen, M., Latonen, L., Ruusuvuori, P., and Marttinen, P. (2023). Deformation equivariant cross-modality image synthesis with paired non-aligned training data. Medical Image Analysis, 90:102940. Available online

• Ji, S. and Marttinen, P. (2023). Patient Outcome and Zero-shot Diagnosis Prediction with Hypernetwork-guided Multitask Learning. In Proceedings of the 17th Conference of the European Chapter of the Association for Computational Linguistics (EACL 2023). Available online

• Marttinen, P. and Hanage, W.P. (2017). Speciation trajectories in recombining bacterial species. PLOS Computational Biology, 13(7):e1005640. Available online

• Mostowy, R., Croucher, N.J., Andam, C.P., Corander, J., Hanage, W.P., and Marttinen, P. (2017). Efficient inference of recent and ancestral recombination within bacterial populations. Molecular Biology and Evolution, 34(5):1167-1182. Available online

• Marttinen, P., Pirinen, M., Sarin, A.P., Gillberg, J., Kettunen, J., Surakka, I., Kangas, A.J., Soininen, P., O’Reilly, P.F., Kaakinen, M., Kähönen, M., Lehtimäki, T., Ala-Korpela, M., Raitakari, O.T., Salomaa, V., Järvelin, M.-R., Ripatti, S. and Kaski, S. (2014). Assessing multivariate gene-metabolome associations with rare variants using Bayesian reduced rank regression. Bioinformatics, 30(14):2026-34. Available online

• Corander, J. and Marttinen, P. (2006). Bayesian identification of admixture events using multi-locus molecular markers. Molecular Ecology, 15:2833-2843. Link

Selected Articles (Applications)

• Ojala, F.*, Abdul Sater, M.R.*, Miller, L.G., McKinnell, J.A., Hayden, M.K., Huang, S.S., Grad, Y.H.+, and Marttinen, P.+ (2023). Bayesian modeling of the impact of antibiotic resistance on the efficiency of MRSA decolonization. PLOS Computational Biology, accepted. (*/+ equal contribution) Preprint

• Chewapreecha, C., Marttinen, P., Croucher, N.J.,Salter, S.J., Harris, S.R., Mather, A.E.,Hanage, W.P., Goldblatt, D., Nosten, F.H., Turner, C., Turner, P., Bentley, S.D. and Parkhill, J. (2014). Comprehensive identification of single nucleotide polymorphisms associated with beta-lactam resistance within pneumococcal mosaic genes. PLoS Genetics, 10(8):e1004547. Available online

• Kashtan, N., Roggensack, S.E., Rodrigue, S., Thompson, J.W., Biller, S.J., Coe, A., Ding, H., Marttinen, P., Malmstrom, R.R., Stocker, R., Follows, M.J., Stepanauskas, R. and Chisholm, S.W. (2014). Single-cell genomics reveals hundreds of coexisting subpopulations in wild Prochlorococcus. Science, 344(6182): 416-420. Link

• Chewapreecha, C., Harris, S.R., Croucher, N.J., Turner, C., Marttinen, P., Cheng, L., Pessia, A., Aanensen, D.M., Mather, A.E., Page, A.J., Salter, S.J., Harris, D., Nosten, F., Goldblatt, D., Corander, J., Parkhill, J., Turner, P. and Bentley, S.D. (2014). Dense genomic sampling identifies highways of pneumococcal recombination. Nature Genetics, 46: 305-309. Link

Software

I have (co-)authored the following software for population genetics. For code related to later articles by the group, see the respective articles.

• BAPS (Examples): Bayesian analysis of genetic population structure, see publications [1,3,4,7] in the full list
• fastGEAR: Inference of recent and ancestral recombination between divergent lineages in bacteria [34]
• BratNextGen (win32, win64, Linux): Bayesian modeling of recombination events in bacterial populations [6,14]
• SEER: A method for genome-wide association studies in bacteria [31]
• BRRR: Bayesian reduced rank regression for multivariate genome-wide association studies [22]
• Bacmeta / cog_sim / habitat_model: Simulation models for microbial population genomics [39,26,37]
• K-Pax: Bayesian clustering of protein sequences and detection of function specific residues [5]

Teaching

I am the responsible professor in the Major of Machine Learning, Data Science and Artificial Intelligence (Macadamia).

Courses I have taught:

• Special course: Seminar on NLP/LLMs: 2024, 2023
• Machine Learning: 2024, 2023, 2022
• Special course: Elements of Causal Inference: 2022, 2021
• Machine Learning: Advanced Probabilistic Methods: 2024, 2023, 2022, 2021, 2020, 2018, 2017, 2016, 2015
• Computational Genomics: 2020, 2019, 2018, 2017 fall, 2017 spring
• Data Science: 2019
• Information visualization: 2015
• Introduction to Bayesian Modeling: 2011, 2010
• Special Course in Bayesian Modeling (Bayesian Networks): 2010