{"type":"ScholarlyArticle","title":"Forming Effective Human-AI Teams: Building Machine Learning Models that Complement the Capabilities of Multiple Experts","authors":[{"name":"Patrick Hemmer","slug":"patrick-hemmer","given":"Patrick","family":"Hemmer","affiliation":"Karlsruhe Institute of Technology"},{"name":"Sebastian Schellhammer","slug":"sebastian-schellhammer","given":"Sebastian","family":"Schellhammer","affiliation":"Karlsruhe Institute of Technology; GESIS -Leibniz Institute for the Social Sciences"},{"name":"Michael Vössing","slug":"michael-vossing","given":"Michael","family":"Vössing","affiliation":"Karlsruhe Institute of Technology"},{"name":"Johannes Jakubik","slug":"johannes-jakubik","given":"Johannes","family":"Jakubik","affiliation":"Karlsruhe Institute of Technology"},{"name":"Gerhard Satzger","slug":"gerhard-satzger","given":"Gerhard","family":"Satzger","affiliation":"Karlsruhe Institute of Technology"}],"abstract":"Machine learning (ML) models are increasingly being used in application domains that often involve working together with human experts. In this context, it can be advantageous to defer certain instances to a single human expert when they are difficult to predict for the ML model. While previous work has focused on scenarios with one distinct human expert, in many real-world situations several human experts with varying capabilities may be available. In this work, we propose an approach that trains a classification model to complement the capabilities of multiple human experts. By jointly training the classifier together with an allocation system, the classifier learns to accurately predict those instances that are difficult for the human experts, while the allocation system learns to pass each instance to the most suitable team member—either the classifier or one of the human experts. We evaluate our proposed approach in multiple experiments on public datasets with “synthetic” experts and a real-world medical dataset annotated by multiple radiologists. Our approach outperforms prior work and is more accurate than the best human expert or a classifier. Furthermore, it is flexibly adaptable to teams of varying sizes and different levels of expert diversity.","doi":"10.24963/ijcai.2022/344","url":"https://ijcai.org/proceedings/2022/344","pdf":"https://ijcai.org/proceedings/2022/0344.pdf","markdown":"https://ijcai.org/proceedings/2022/344.md","bibtex":"https://ijcai.org/proceedings/2022/bibtex/344","volume":{"slug":"2022","title":"Proceedings of the Thirty-First International Joint Conference on Artificial Intelligence","short_title":"IJCAI 2022","url":"https://ijcai.org/proceedings/2022/","year":2022,"location":"Vienna","dates":"23-29 July 2022","editors":"Luc De Raedt","isbn":"978-1-956792-00-3","publisher":"International Joint Conferences on Artificial Intelligence Organization","doi_prefix":"10.24963/ijcai.2022"},"section":"Humans and AI","track":"Main Track","number":344,"pages":{"first":2478,"last":2484},"keywords":["Humans and AI: Human-AI Collaboration"],"videos":[{"label":"Video #1","url":"https://slideslive.com/38987121","length_sec":88,"provider":"slideslive"},{"label":"Video #2","url":"https://slideslive.com/38985264","length_sec":711,"provider":"slideslive"}],"references":[{"position":1,"raw":"[Bansal et al., 2021] Gagan Bansal, Besmira Nushi, Ece Ka- mar, Eric Horvitz, and Daniel S. Weld. Is the most accu- rate ai the best teammate? optimizing ai for teamwork. In AAAI, 2021.","title":"Is the most accurate ai the best teammate? optimizing ai for teamwork","authors":"Gagan Bansal, Besmira Nushi, Ece Kamar, Eric Horvitz, Daniel S Weld","year":"2021","venue":"AAAI","doi":null,"url":null,"internal_url":null},{"position":2,"raw":"[Chouldechova et al., 2018] Alexandra Chouldechova, Di- ana Benavides Prado, Oleksandr Fialko, and Rhema Vaithianathan. A case study of algorithm-assisted decision making in child maltreatment hotline screening decisions. In FAT, 2018.","title":"A case study of algorithm-assisted decision making in child maltreatment hotline screening decisions","authors":"Alexandra Chouldechova, Diana Benavides Prado, Oleksandr Fialko, Rhema Vaithianathan","year":"2018","venue":"FAT","doi":null,"url":null,"internal_url":null},{"position":3,"raw":"[Davidson et al., 2017] Thomas Davidson, Dana Warmsley, Michael W. Macy, and Ingmar Weber. Automated hate speech detection and the problem of offensive language. In ICWSM, 2017.","title":"Automated hate speech detection and the problem of offensive language","authors":"Thomas Davidson, Dana Warmsley, Michael W Macy, Ingmar Weber","year":"2017","venue":"ICWSM","doi":null,"url":null,"internal_url":null},{"position":4,"raw":"[De et al., 2021] Abir De, Nastaran Okati, Ali Zarezade, and Manuel Gomez-Rodriguez. Classification under human assistance. In AAAI, 2021.","title":"Classification under human assistance","authors":"Abir De, Nastaran Okati, Ali Zarezade, Manuel Gomez-Rodriguez","year":"2021","venue":"AAAI","doi":null,"url":null,"internal_url":null},{"position":5,"raw":"[Esteva et al., 2017] Andre Esteva, Brett Kuprel, Roberto A. Novoa, Justin M. Ko, Susan M. Swetter, Helen M. Blau, and Sebastian Thrun. Dermatologist-level classification of skin cancer with deep neural networks. Nature, 2017.","title":"Dermatologist-level classification of skin cancer with deep neural networks","authors":"Andre Esteva, Brett Kuprel, Roberto A Novoa, Justin M Ko, Susan M Swetter, Helen M Blau, Sebastian Thrun","year":"2017","venue":"Nature","doi":null,"url":null,"internal_url":null},{"position":6,"raw":"[Gao et al., 2021] Ruijiang Gao, Maytal Saar-Tsechansky, Maria De-Arteaga, Ligong Han, Min Kyung Lee, and Matthew Lease. Human-ai collaboration with bandit feed- back. In IJCAI, 2021.","title":"Human-ai collaboration with bandit feedback","authors":"Gao","year":"2021","venue":"IJCAI","doi":null,"url":null,"internal_url":"/proceedings/2021/237"},{"position":7,"raw":"[Grønsund and Aanestad, 2020] Tor Grønsund and Margunn Aanestad. Augmenting the algorithm: emerging human- in-the-loop work configurations. The Journal of Strategic Information Systems, 2020.","title":"Tor Grønsund and Margunn Aanestad. Augmenting the algorithm: emerging humanin-the-loop work configurations","authors":"Aanestad Grønsund","year":"2020","venue":"The Journal of Strategic Information Systems","doi":null,"url":null,"internal_url":null},{"position":8,"raw":"[Gulshan et al., 2019] Varun Gulshan, Renu P. Rajan, Ka- sumi Widner, Derek J. Wu, Peter Wubbels, Tyler Rhodes, Kira Whitehouse, Marc Coram, Greg S Corrado, Kim Ra- masamy, Rajiv Raman, Lily H. Peng, and Dale R. Web- ster. Performance of a deep-learning algorithm vs manual grading for detecting diabetic retinopathy in india. JAMA Ophthalmology, 2019.","title":"Performance of a deep-learning algorithm vs manual grading for detecting diabetic retinopathy in india","authors":"Varun Gulshan, P Renu, Kasumi Rajan, Derek J Widner, Peter Wu, Tyler Wubbels, Kira Rhodes, Marc Whitehouse, Greg S Coram, Kim Corrado, Rajiv Ramasamy, Lily H Raman, Dale R Peng","year":"2019","venue":"JAMA Ophthalmology","doi":null,"url":null,"internal_url":null},{"position":9,"raw":"[He et al., 2016] Kaiming He, Xiangyu Zhang, Shaoqing Ren, and Jian Sun. Deep residual learning for image recog- nition. In CVPR, 2016.","title":"Deep residual learning for image recognition","authors":"Kaiming He, Xiangyu Zhang, Shaoqing Ren, Jian Sun","year":"2016","venue":"CVPR","doi":null,"url":null,"internal_url":null},{"position":10,"raw":"[Hemmer et al., 2021] Patrick Hemmer, Max Schemmer, Michael Vössing, and Niklas Kühl. Human-ai comple- mentarity in hybrid intelligence systems: a structured lit- erature review. In PACIS, 2021.","title":"Human-ai complementarity in hybrid intelligence systems: a structured literature review","authors":"Patrick Hemmer, Max Schemmer, Michael Vössing, Niklas Kühl","year":"2021","venue":"PACIS","doi":null,"url":null,"internal_url":null},{"position":11,"raw":"[Irvin et al., 2019] Jeremy Irvin, Pranav Rajpurkar, Michael Ko, Yifan Yu, Silviana Ciurea-Ilcus, Chris Chute, Henrik Marklund, Behzad Haghgoo, Robyn Ball, et al. Chexpert: a large chest radiograph dataset with uncertainty labels and expert comparison. In AAAI, 2019.","title":"Chexpert: a large chest radiograph dataset with uncertainty labels and expert comparison","authors":"Jeremy Irvin, Pranav Rajpurkar, Michael Ko, Yifan Yu, Silviana Ciurea-Ilcus, Chris Chute, Henrik Marklund, Behzad Haghgoo, Robyn Ball","year":"2019","venue":"AAAI","doi":null,"url":null,"internal_url":null},{"position":12,"raw":"[Jacobs et al., 1991] Robert A. Jacobs, Michael I. Jordan, Steven J. Nowlan, and Geoffrey E. Hinton. Adaptive mix- tures of local experts. Neural Computation, 1991.","title":"Adaptive mixtures of local experts","authors":"Robert A Jacobs, Michael I Jordan, Steven J Nowlan, Geoffrey E Hinton","year":"1991","venue":"Neural Computation","doi":null,"url":null,"internal_url":null},{"position":13,"raw":"[Keswani et al., 2021] Vijay Keswani, Matthew Lease, and Krishnaram Kenthapadi. Towards unbiased and accurate deferral to multiple experts. In AIES, 2021. [Krizhevsky, 2009] Alex Krizhevsky. Learning multiple lay- ers of features from tiny images. Technical report, Univer- sity of Toronto, 2009.","title":"Towards unbiased and accurate deferral to multiple experts","authors":"Vijay Keswani, Matthew Lease, Krishnaram Kenthapadi","year":"2009","venue":"Alex Krizhevsky. Learning multiple layers of features from tiny images","doi":null,"url":null,"internal_url":null},{"position":14,"raw":"[Lampert et al., 2016] Thomas A Lampert, André Stumpf, and Pierre Ganc ¸arski. An empirical study into annotator agreement, ground truth estimation, and algorithm evalua- tion. IEEE Transactions on Image Processing, 2016.","title":"An empirical study into annotator agreement, ground truth estimation, and algorithm evaluation","authors":"André Thomas A Lampert, Pierre Stumpf, Ganc ¸arski","year":"2016","venue":"IEEE Transactions on Image Processing","doi":null,"url":null,"internal_url":null},{"position":15,"raw":"[Lintott et al., 2008] Chris J. Lintott, Kevin Schawinski, Anže Slosar, Kate R. Land, Steven Bamford, Daniel I. Thomas, M. Jordan Raddick, Robert Nichol, Alexander S. Szalay, Daniel Andreescu, P. G. Murray, and Jan van den Berg. Galaxy zoo: morphologies derived from visual inspection of galaxies from the sloan digital sky survey. Monthly Notices of the Royal Astronomical Society, 2008.","title":"Galaxy zoo: morphologies derived from visual inspection of galaxies from the sloan digital sky survey","authors":"Chris J Lintott, Kevin Schawinski, Anže Slosar, Kate R Land, Steven Bamford, Daniel I Thomas, M Jordan Raddick, Robert Nichol, Alexander S Szalay, P G Daniel Andreescu, Jan Murray, Van Den, Berg","year":"2008","venue":"Monthly Notices of the Royal Astronomical Society","doi":null,"url":null,"internal_url":null},{"position":16,"raw":"[Madras et al., 2018] David Madras, Toniann Pitassi, and Richard S. Zemel. Predict responsibly: improving fairness and accuracy by learning to defer. In NeurIPS, 2018.","title":"Predict responsibly: improving fairness and accuracy by learning to defer","authors":"David Madras, Toniann Pitassi, Richard S Zemel","year":"2018","venue":"NeurIPS","doi":null,"url":null,"internal_url":null},{"position":17,"raw":"[Majkowska et al., 2020] Anna Majkowska, Sid Mittal, David F Steiner, Joshua J Reicher, Scott Mayer McK- inney, Gavin E Duggan, Krish Eswaran, Po-Hsuan Cameron Chen, Yun Liu, Kalidindi, et al. Chest radio- graph interpretation with deep learning models: assess- ment with radiologist-adjudicated reference standards and population-adjusted evaluation. Radiology, 2020.","title":"Chest radiograph interpretation with deep learning models: assessment with radiologist-adjudicated reference standards and population-adjusted evaluation","authors":"Anna Majkowska, Sid Mittal, David F Steiner, Joshua J Reicher, Scott Mayer Mckinney, Gavin E Duggan, Krish Eswaran, Yun Po-Hsuan Cameron Chen, Liu, Kalidindi","year":"2020","venue":"Radiology","doi":null,"url":null,"internal_url":null},{"position":18,"raw":"[McInnes et al., 2018] Leland McInnes, John Healy, and James Melville. Umap: uniform manifold approximation and projection for dimension reduction. arXiv preprint arXiv:1802.03426, 2018.","title":"Umap: uniform manifold approximation and projection for dimension reduction","authors":"Leland Mcinnes, John Healy, James Melville","year":"2018","venue":null,"doi":null,"url":null,"internal_url":null},{"position":19,"raw":"[Mozannar and Sontag, 2020] Hussein Mozannar and David A. Sontag. Consistent estimators for learning to defer to an expert. In ICML, 2020.","title":"Consistent estimators for learning to defer to an expert","authors":"Hussein Mozannar, David A Sontag","year":"2020","venue":"ICML","doi":null,"url":null,"internal_url":null},{"position":20,"raw":"[Okati et al., 2021] Nastaran Okati, Abir De, and Manuel Rodriguez. Differentiable learning under triage. In NeurIPS, 2021.","title":"Differentiable learning under triage","authors":"Nastaran Okati, Abir De, Manuel Rodriguez","year":"2021","venue":"NeurIPS","doi":null,"url":null,"internal_url":null},{"position":21,"raw":"[Pennington et al., 2014] Jeffrey Pennington, Richard Socher, and Christopher D. Manning. Glove: global vectors for word representation. In EMNLP, 2014.","title":"Glove: global vectors for word representation","authors":"Jeffrey Pennington, Richard Socher, Christopher D Manning","year":"2014","venue":"EMNLP","doi":null,"url":null,"internal_url":null},{"position":22,"raw":"[Peterson et al., 2019] Joshua C. Peterson, Ruairidh M. Bat- tleday, Thomas L. Griffiths, and Olga Russakovsky. Hu- man uncertainty makes classification more robust. In ICCV, 2019.","title":"Human uncertainty makes classification more robust","authors":"Joshua C Peterson, Ruairidh M Battleday, Thomas L Griffiths, Olga Russakovsky","year":"2019","venue":"ICCV","doi":null,"url":null,"internal_url":null},{"position":23,"raw":"[Raghu et al., 2019] Maithra Raghu, Katy Blumer, Greg Corrado, Jon Kleinberg, Ziad Obermeyer, and Send- hil Mullainathan. The algorithmic automation problem: prediction, triage, and human effort. arXiv preprint arXiv:1903.12220, 2019.","title":"The algorithmic automation problem: prediction, triage, and human effort","authors":"Raghu","year":"2019","venue":null,"doi":null,"url":null,"internal_url":null},{"position":24,"raw":"[Wang et al., 2017] Xiaosong Wang, Yifan Peng, Le Lu, Zhiyong Lu, Mohammadhadi Bagheri, and Ronald M. Summers. Chestx-ray8: hospital-scale chest x-ray database and benchmarks on weakly-supervised classifica- tion and localization of common thorax diseases. In CVPR, 2017.","title":"Chestx-ray8: hospital-scale chest x-ray database and benchmarks on weakly-supervised classification and localization of common thorax diseases","authors":"Xiaosong Wang, Yifan Peng, Le Lu, Zhiyong Lu, Mohammadhadi Bagheri, Ronald M Summers","year":"2017","venue":"CVPR","doi":null,"url":null,"internal_url":null},{"position":25,"raw":"[Wilder et al., 2020] Bryan Wilder, Eric Horvitz, and Ece Kamar. Learning to complement humans. In IJCAI, 2020.","title":"Learning to complement humans","authors":"Bryan Wilder, Eric Horvitz, Ece Kamar","year":"2020","venue":"IJCAI","doi":null,"url":null,"internal_url":"/proceedings/2020/212"}],"license_note":"Copyright International Joint Conferences on Artificial Intelligence."}