Fredrik D. Johansson, PhD
Associate professor of Computer Science & Engineering
and PI for the Healthy AI Lab
at Chalmers University of Technology, Sweden
Machine Learning for Decision Making & Healthcare, CV (May, 2024)
Research
My research interests are primarily focused on causal inference in machine learning and decision-making with applications in healthcare.
Current PhD students (as main advisor):
- Herman Bergström
- Alessandro Margueritte
- Ahmet Balcioglu
- Adam Breitholtz
- Newton Mwai Kinyanjui
- Anton Matsson
- Lena Stempfle
I was previously a post-doctoral associate in the Clinical Machine Learning group at CSAIL & IMES, Massachusetts Institute of Technology working with Professor David Sontag.
Contact
You can contact me by email at fredrik.johansson@chalmers.se or ml@fredjo.com.
Publications
The list below is no longer up to date. Here is my Google Scholar profile.Machine learning methodology & theory, causality
- Fredrik D. Johansson IncomeSCM: From tabular data set to time-series simulator and causal estimation benchmark. Advances in Neural Information Processing Systems (To appear). 2024
- Herman Bergström, Emil Carlsson, Devdatt Dubhashi, Fredrik D. Johansson Active Preference Learning for Ordering Items In-and Out-of-sample. Advances in Neural Information Processing Systems (To appear). 2024
- Lena Stempfle, Fredrik D. Johansson MINTY: Rule-based models that minimize the need for imputing features with missing values. International Conference on Artificial Intelligence and Statistics (AISTATS). 2024
- Emil Carlsson, Debabrota Basu, Fredrik D. Johansson, Devdatt Dubhashi Pure exploration in bandits with linear constraints. International Conference on Artificial Intelligence and Statistics (AISTATS). 2024
- Sofia Ek, Dave Zachariah, Fredrik D. Johansson, Peter Stoica Off-Policy Evaluation with Out-of-Sample Guarantees. Transactions on Machine Learning Research (TMLR). 2023
- Connor T. Jerzak, Fredrik D. Johansson, Adel Daoud Image-based Treatment Effect Heterogeneity. Caual Learning and Reasoning (CLeaR). 2023
- Lena Stempfle, Ashkan Panahi, Fredrik D. Johansson Sharing pattern submodels for prediction with missing values. AAAI Conference on Artificial Intelligence (AAAI). 2023.
- Efficient learning of nonlinear prediction models with time-series privileged information Efficient learning of nonlinear prediction models with time-series privileged information. Neural Information Processing Systems (NeurIPS). 2022.
- Adam Breitholtz, Fredrik D. Johansson Practicality of generalization guarantees for unsupervised domain adaptation with neural networks. Transactions on Machine Learning Research (TMLR). 2022.
- Fredrik D. Johansson, Uri Shalit, Nathan Kallus, David Sontag. Generalization Bounds and Representation Learning for Estimation of Potential Outcomes and Causal Effects. Journal of Machine Learning Research (JMLR). 2022.
- Anton Matsson, Fredrik D. Johansson, Case-based off-policy policy evaluation using prototype learning. Proceedings of Uncertainty in Artificial Intelligence (UAI). 2022.
- Newton Mwai Kinyanjui, Fredrik D. Johansson. ADCB: An Alzheimer's disease simulator for benchmarking observational estimators of causal effects. Proceedings of Conference on Health, Inference, and Learning, 2022.
- Rickard Karlsson, Martin Willbo, Zeshan Hussain, Rahul G. Krishnan, David Sontag, Fredrik D. Johansson Using Time-Series Privileged Information for Provably Efficient Learning of Prediction Models. AISTATS, 2022.
- Emil Carlsson, Devdatt Dubhashi, Fredrik D. Johansson Thompson Sampling for Bandits with Clustered Arms. IJCAI, 2021.
- Emil Carlsson, Devdatt Dubhashi, Fredrik D. Johansson Learning Approximate and Exact Numeral Systems via Reinforcement Learning. Proceedings of the Annual Meeting of the Cognitive Science Society, CogSci, 2021.
- Nils M Kriege, Fredrik D Johansson, Christopher Morris A Survey on Graph Kernels. Applied Network Science, 5 (1), 2020.
- Samuel Håkansson, Viktor Lindblom, Omer Gottesman, Fredrik D. Johansson. Learning to search efficiently for causally near-optimal treatments. NeurIPS, 2020.
- Maggie Makar, Fredrik D. Johansson, John Guttag, David Sontag Estimation of Bounds on Potential Outcomes For Decision Making. ICML, 2020.
- Fredrik D. Johansson, Dennis Wei, Michael Oberst, Tian Gao, Gabriel Brat, David Sontag, Kush R. Varshney Characterization of Overlap in Observational Studies. AISTATS, 2020.
- Fredrik D. Johansson, Uri Shalit, Nathan Kallus, David Sontag Generalization Bounds and Representation Learning for Estimation of Potential Outcomes and Causal Effects. arXiv preprint arXiv:2001.07426, 2020.
- Fredrik D. Johansson, David Sontag, Rajesh Ranganath. Support and Invertibility in Domain-Invariant Representations. In Proc. of the 22nd International Conference on Artificial Intelligence and Statistics, AISTATS, 2019.
- Irene Chen, Fredrik D. Johansson, David Sontag. Why is my classifier discriminatory? In Proc. of Neural Information Processing Systems, NIPS, 2018.
- Ashkan Panahi, Devdatt Dubhashi, Fredrik Johansson, Chiranjib Bhattacharyya Clustering by Sum of Norms: Stochastic Incremental Algorithm, Convergence and Cluster Recovery. ICML 2017.
-
Uri Shalit*, Fredrik D. Johansson*, David Sontag.
Estimating individual treatment effect: generalization bounds and algorithms. ICML 2017. [Code] [Data links posted below]
*Equal contribution -
Fredrik D. Johansson*, Uri Shalit*, David Sontag.
Learning Representations for Counterfactual Inference.
ICML 2016 [Slides]
[Code]
*Equal contribution - Fredrik D. Johansson, Ankani Chattoraj, Chiranjib Bhattacharyya, Devdatt Dubhashi. Weighted Theta Functions and Embeddings with Applications to Max-Cut, Clustering and Summarization. NIPS 2015
- Fredrik D. Johansson, Otto Frost, Carl Retzner, and Devdatt Dubhashi. Classifying large graphs with differential privacy. MDAI 2015
- Linus Hermansson, Fredrik D. Johansson and Osamu Watanabe. Generalized Shortest Path Kernel on Graphs. Discovery Science 2015
- Fredrik Johansson, Vinay Jethava, Devdatt Dubhashi, Chiranjib Bhattacharyya. Learning with similarity functions on graphs using matchings of geometric embeddings. KDD 2015
- Fredrik Axelsson, Birger Rydback, Fredrik Johansson, Jonathan Bengtsson, Svetoslav Marinov Data-driven Coreference Resolution for Swedish. SLTC 2014
- Fredrik Johansson, Vinay Jethava, Devdatt Dubhashi, Chiranjib Bhattacharyya. Global graph kernels using geometric embeddings. ICML 2013
- Tommi Kerola, Linus Hermansson, Fredrik Johansson, Vinay Jethava, Devdatt Dubhashi. Entity Disambiguation in Anonymized Graphs Using Graph Kernels. CIKM 2013
Healthcare & other applications
- Hákon Valur Dansson, Lena Stempfle, Hildur Egilsdóttir, Alexander Schliep; Erik Portelius, Kaj Blennow, Henrik Zetterberg, Fredrik D. Johansson Predicting progression & cognitive decline in amyloid-positive patients with Alzheimer's disease. Accepted for publication in Alzheimer's Research and Therapy, 2021.
- Fredrik D Johansson, Jamie E Collins, Vincent Yau, Hongshu Guan, Seoyoung C Kim, Elena Losina, David Sontag, Jacklyn Stratton, Huong Trinh, Jeffrey Greenberg, Daniel H Solomon Predicting Response to Tocilizumab Monotherapy in Rheumatoid Arthritis: A Real-World Data Analysis Using Machine Learning. The Journal of Rheumatology, 2021
- Daniel H Solomon, Chang Xu, Jamie Collins, Seoyoung C Kim, Elena Losina, Vincent Yau, Fredrik D Johansson The sequence of disease-modifying anti-rheumatic drugs: pathways to and predictors of tocilizumab monotherapy. Arthritis Research & Therapy 23 (1), 1--9, 2021
- Jamie E Collins, Fredrik D Johansson, Sara Gale, Seoyoung Kim, Swastina Shrestha, David Sontag, Jacklyn Stratton, Huong Trinh, Chang Xu, Elena Losina, Daniel H Solomon Predicting Remission Among Patients With Rheumatoid Arthritis Starting Tocilizumab Monotherapy: Model Derivation and Remission Score Development. ACR open rheumatology 2 (2), 65--73, 2020
- Fredrik D Johansson Machine Learning Analysis of Heterogeneity in the Effect of Student Mindset Interventions. Observational Studies, 2019.
- Adel Daoud, Fredrik D Johansson Estimating Treatment Heterogeneity of International Monetary Fund Programs on Child Poverty with Generalized Random Forest. socarxiv preprint, 2019.
- Omer Gottesman, Fredrik Johansson, Matthieu Komorowski, Aldo Faisal, David Sontag, Finale Doshi-Velez, Leo Anthony Celi Guidelines for reinforcement learning in healthcare. Nature Medicine 25 (1), 16--18, 2019
- Nina Tahmasebi, Lars Borin, Gabriele Capannini, Devdatt Dubhashi, Peter Exner, Markus Forsberg, Gerhard Gossen, Fredrik D Johansson, Richard Johansson, Mikael Kågebäck, Olof Mogren, Pierre Nugues, Thomas Risse. Visions and open challenges for a knowledge-based culturomics. Int J Digit Libr (2015) 15: 169, 2015
Technical reports \& workshop papers
- Omer Gottesman, Fredrik D. Johansson, et al. Evaluating Reinforcement Learning Algorithms in Observational Health Settings. In arXiv preprint arXiv:1805.12298, 2018.
- Fredrik D. Johansson, Nathan Kallus, Uri Shalit, David Sontag. Learning Weighted Representations for Generalization Across Designs. In arXiv preprint arXiv:1802.08598, 2018.
- Emilio Jorge, Mikael Kågebäck, Fredrik D. Johansson, Emil Gustavsson. Learning to Play Guess Who? and Inventing a Grounded Language as a Consequence. In arXiv preprint arXiv:1611.03218, 2016.
- Mikael Kågebäck, Fredrik Johansson, Richard Johansson, Devdatt Dubhashi. Neural context embeddings for automatic discovery of word senses. NAACL-HLT 2015
- Fredrik Johansson, Vinay Jethava, Devdatt Dubhashi. DLOREAN: Dynamic Location-Aware Reconstruction of Multiway Networks. ICDMW 2013
- Fredrik Johansson, Tobias Färdig, Vinay Jethava, Svetoslav Marinov. Intent-aware temporal query modeling for keyword suggestion. PIKM 2012 p.83-86
PhD Thesis
- Fredrik Johansson. Learning with geometric embeddings of graphs . Doctoral thesis. 2016
Software & Data
Counterfactual Regression is a package for treatment effect estimation in Python.
IHDP-100 (train), IHDP-100 (test) are 100 realizations of the IHDP dataset as used in Shalit, J, Sontag, ICML, 2017. Variables in the .npz x, t, yf, ycf, mu0, mu1 are covariates, treatment, factual outcome, counterfactual outcome, and noiseless potential outcomes respectively.
Jobs-Binary (train), Jobs-Binary (test) is the preprocessed Jobs dataset, following the same format as IHDP, as used in Shalit, J, Sontag, ICML, 2017. Variables in the .npz x, t, yf, e are covariates, treatment, factual outcome and indicator for original randomized sample respectively.
IHDP-1000 (train), IHDP-1000 (test) are 1000 realizations of the IHDP dataset as used in Shalit, J, Sontag, ICML, 2017.
News, is a dataset for treatment effect estimation as used in J, Shalit, Sontag, ICML, 2016.
Matching Kernels is a package of graph kernels for MATLAB.
Lovasz-theta Kernel is a package of graph kernels for MATLAB.
DLOREAN is a software for inferring multi-way dynamic networks from node data.
Entity Graphs is a collection of graphs of entity interactions built from data collected by Recorded Future.