Acquisition of chess knowledge in AlphaZero McGrath, Thomas; Kapishnikov, Andrei; Tomašev, Nenad ...
Proceedings of the National Academy of Sciences - PNAS,
11/2022, Letnik:
119, Številka:
47
Journal Article
Recenzirano
Odprti dostop
We analyze the knowledge acquired by AlphaZero, a neural network engine that learns chess solely by playing against itself yet becomes capable of outperforming human chess players. Although the ...system trains without access to human games or guidance, it appears to learn concepts analogous to those used by human chess players. We provide two lines of evidence. Linear probes applied to AlphaZero's internal state enable us to quantify when and where such concepts are represented in the network. We also describe a behavioral analysis of opening play, including qualitative commentary by a former world chess champion.
The practice of mathematics involves discovering patterns and using these to formulate and prove conjectures, resulting in theorems. Since the 1960s, mathematicians have used computers to assist in ...the discovery of patterns and formulation of conjectures
, most famously in the Birch and Swinnerton-Dyer conjecture
, a Millennium Prize Problem
. Here we provide examples of new fundamental results in pure mathematics that have been discovered with the assistance of machine learning-demonstrating a method by which machine learning can aid mathematicians in discovering new conjectures and theorems. We propose a process of using machine learning to discover potential patterns and relations between mathematical objects, understanding them with attribution techniques and using these observations to guide intuition and propose conjectures. We outline this machine-learning-guided framework and demonstrate its successful application to current research questions in distinct areas of pure mathematics, in each case showing how it led to meaningful mathematical contributions on important open problems: a new connection between the algebraic and geometric structure of knots, and a candidate algorithm predicted by the combinatorial invariance conjecture for symmetric groups
. Our work may serve as a model for collaboration between the fields of mathematics and artificial intelligence (AI) that can achieve surprising results by leveraging the respective strengths of mathematicians and machine learning.
The volume and complexity of diagnostic imaging is increasing at a pace faster than the availability of human expertise to interpret it. Artificial intelligence has shown great promise in classifying ...two-dimensional photographs of some common diseases and typically relies on databases of millions of annotated images. Until now, the challenge of reaching the performance of expert clinicians in a real-world clinical pathway with three-dimensional diagnostic scans has remained unsolved. Here, we apply a novel deep learning architecture to a clinically heterogeneous set of three-dimensional optical coherence tomography scans from patients referred to a major eye hospital. We demonstrate performance in making a referral recommendation that reaches or exceeds that of experts on a range of sight-threatening retinal diseases after training on only 14,884 scans. Moreover, we demonstrate that the tissue segmentations produced by our architecture act as a device-independent representation; referral accuracy is maintained when using tissue segmentations from a different type of device. Our work removes previous barriers to wider clinical use without prohibitive training data requirements across multiple pathologies in a real-world setting.
The Role of Hubness in Clustering High-Dimensional Data Tomasev, Nenad; Radovanovic, Milos; Mladenic, Dunja ...
IEEE transactions on knowledge and data engineering,
2014-March, 2014-03-00, 20140301, Letnik:
26, Številka:
3
Journal Article
Recenzirano
Odprti dostop
High-dimensional data arise naturally in many domains, and have regularly presented a great challenge for traditional data mining techniques, both in terms of effectiveness and efficiency. Clustering ...becomes difficult due to the increasing sparsity of such data, as well as the increasing difficulty in distinguishing distances between data points. In this paper, we take a novel perspective on the problem of clustering high-dimensional data. Instead of attempting to avoid the curse of dimensionality by observing a lower dimensional feature subspace, we embrace dimensionality by taking advantage of inherently high-dimensional phenomena. More specifically, we show that hubness, i.e., the tendency of high-dimensional data to contain points (hubs) that frequently occur in k-nearest-neighbor lists of other points, can be successfully exploited in clustering. We validate our hypothesis by demonstrating that hubness is a good measure of point centrality within a high-dimensional data cluster, and by proposing several hubness-based clustering algorithms, showing that major hubs can be used effectively as cluster prototypes or as guides during the search for centroid-based cluster configurations. Experimental results demonstrate good performance of our algorithms in multiple settings, particularly in the presence of large quantities of noise. The proposed methods are tailored mostly for detecting approximately hyperspherical clusters and need to be extended to properly handle clusters of arbitrary shapes.
Machine learning (ML) holds great promise for improving healthcare, but it is critical to ensure that its use will not propagate or amplify health disparities. An important step is to characterize ...the (un)fairness of ML models-their tendency to perform differently across subgroups of the population-and to understand its underlying mechanisms. One potential driver of algorithmic unfairness, shortcut learning, arises when ML models base predictions on improper correlations in the training data. Diagnosing this phenomenon is difficult as sensitive attributes may be causally linked with disease. Using multitask learning, we propose a method to directly test for the presence of shortcut learning in clinical ML systems and demonstrate its application to clinical tasks in radiology and dermatology. Finally, our approach reveals instances when shortcutting is not responsible for unfairness, highlighting the need for a holistic approach to fairness mitigation in medical AI.
Learning with label noise is an important issue in classification, since it is not always possible to obtain reliable data labels. In this paper we explore and evaluate a new approach to learning ...with label noise in intrinsically high-dimensional data, based on using neighbor occurrence models for hubness-aware k-nearest neighbor classification. Hubness is an important aspect of the curse of dimensionality that has a negative effect on many types of similarity-based learning methods. As we will show, the emergence of hubs as centers of influence in high-dimensional data affects the learning process in the presence of label noise. We evaluate the potential impact of hub-centered noise by defining a hubness-proportional random label noise model that is shown to induce a significantly higher kNN misclassification rate than the uniform random label noise. Real-world examples are discussed where hubness-correlated noise arises either naturally or as a consequence of an adversarial attack. Our experimental evaluation reveals that hubness-based fuzzy k-nearest neighbor classification and Naive Hubness-Bayesian k-nearest neighbor classification might be suitable for learning under label noise in intrinsically high-dimensional data, as they exhibit robustness to high levels of random label noise and hubness-proportional random label noise. The results demonstrate promising performance across several data domains.
In modern information societies, there are information systems that track and log parts of the ongoing political discourse. Due to the sheer volume of the accumulated data, automated tools are ...required in order to enable citizens to better interpret political statements and promises, as well as evaluate their truthfulness. We propose an approach to use the established machine learning and data mining techniques for analyzing annotated political statements and promises available via the Serbian Truth-o-meter (Istinomer) system in order to extract and interpret the hidden patterns of truthfulness and deceit. We perform standard textual processing and topic extraction and associate topical truthfulness profiles with the promise makers, for pattern discovery and prediction. Prevailing trends in Serbian political discourse emerge as strong association rules where truthfulness is set as the target variable. The evaluated set of standard content-based prediction models exhibit a bias towards the negative outcomes, due to an overall low truthfulness rate in the data. Our results demonstrate that it is possible to use data mining within political information systems for generating insights into the workings of governments.
IntroductionStandards for Reporting of Diagnostic Accuracy Study (STARD) was developed to improve the completeness and transparency of reporting in studies investigating diagnostic test accuracy. ...However, its current form, STARD 2015 does not address the issues and challenges raised by artificial intelligence (AI)-centred interventions. As such, we propose an AI-specific version of the STARD checklist (STARD-AI), which focuses on the reporting of AI diagnostic test accuracy studies. This paper describes the methods that will be used to develop STARD-AI.Methods and analysisThe development of the STARD-AI checklist can be distilled into six stages. (1) A project organisation phase has been undertaken, during which a Project Team and a Steering Committee were established; (2) An item generation process has been completed following a literature review, a patient and public involvement and engagement exercise and an online scoping survey of international experts; (3) A three-round modified Delphi consensus methodology is underway, which will culminate in a teleconference consensus meeting of experts; (4) Thereafter, the Project Team will draft the initial STARD-AI checklist and the accompanying documents; (5) A piloting phase among expert users will be undertaken to identify items which are either unclear or missing. This process, consisting of surveys and semistructured interviews, will contribute towards the explanation and elaboration document and (6) On finalisation of the manuscripts, the group’s efforts turn towards an organised dissemination and implementation strategy to maximise end-user adoption.Ethics and disseminationEthical approval has been granted by the Joint Research Compliance Office at Imperial College London (reference number: 19IC5679). A dissemination strategy will be aimed towards five groups of stakeholders: (1) academia, (2) policy, (3) guidelines and regulation, (4) industry and (5) public and non-specific stakeholders. We anticipate that dissemination will take place in Q3 of 2021.
Most machine learning tasks involve learning from high-dimensional data, which is often quite difficult to handle. Hubness is an aspect of the curse of dimensionality that was shown to be highly ...detrimental to k-nearest neighbor methods in high-dimensional feature spaces. Hubs, very frequent nearest neighbors, emerge as centers of influence within the data and often act as semantic singularities. This paper deals with evaluating the impact of hubness on learning under class imbalance with k-nearest neighbor methods. Our results suggest that, contrary to the common belief, minority class hubs might be responsible for most misclassification in many high-dimensional datasets. The standard approaches to learning under class imbalance usually clearly favor the instances of the minority class and are not well suited for handling such highly detrimental minority points. In our experiments, we have evaluated several state-of-the-art hubness-aware kNN classifiers that are based on learning from the neighbor occurrence models calculated from the training data. The experiments included learning under severe class imbalance, class overlap and mislabeling and the results suggest that the hubness-aware methods usually achieve promising results on the examined high-dimensional datasets. The improvements seem to be most pronounced when handling the difficult point types: borderline points, rare points and outliers. On most examined datasets, the hubness-aware approaches improve the classification precision of the minority classes and the recall of the majority class, which helps with reducing the negative impact of minority hubs. We argue that it might prove beneficial to combine the extensible hubness-aware voting frameworks with the existing class imbalanced kNN classifiers, in order to properly handle class imbalanced data in high-dimensional feature spaces.
IntroductionFor artificial intelligence (AI) to help improve mental healthcare, the design of data-driven technologies needs to be fair, safe, and inclusive. Participatory design can play a critical ...role in empowering marginalised communities to take an active role in constructing research agendas and outputs. Given the unmet needs of the LGBTQI+ (Lesbian, Gay, Bisexual, Transgender, Queer and Intersex) community in mental healthcare, there is a pressing need for participatory research to include a range of diverse queer perspectives on issues of data collection and use (in routine clinical care as well as for research) as well as AI design. Here we propose a protocol for a Delphi consensus process for the development of PARticipatory Queer AI Research for Mental Health (PARQAIR-MH) practices, aimed at informing digital health practices and policy.Methods and analysisThe development of PARQAIR-MH is comprised of four stages. In stage 1, a review of recent literature and fact-finding consultation with stakeholder organisations will be conducted to define a terms-of-reference for stage 2, the Delphi process. Our Delphi process consists of three rounds, where the first two rounds will iterate and identify items to be included in the final Delphi survey for consensus ratings. Stage 3 consists of consensus meetings to review and aggregate the Delphi survey responses, leading to stage 4 where we will produce a reusable toolkit to facilitate participatory development of future bespoke LGBTQI+–adapted data collection, harmonisation, and use for data-driven AI applications specifically in mental healthcare settings.Ethics and disseminationPARQAIR-MH aims to deliver a toolkit that will help to ensure that the specific needs of LGBTQI+ communities are accounted for in mental health applications of data-driven technologies. The study is expected to run from June 2024 through January 2025, with the final outputs delivered in mid-2025. Participants in the Delphi process will be recruited by snowball and opportunistic sampling via professional networks and social media (but not by direct approach to healthcare service users, patients, specific clinical services, or via clinicians’ caseloads). Participants will not be required to share personal narratives and experiences of healthcare or treatment for any condition. Before agreeing to participate, people will be given information about the issues considered to be in-scope for the Delphi (eg, developing best practices and methods for collecting and harmonising sensitive characteristics data; developing guidelines for data use/reuse) alongside specific risks of unintended harm from participating that can be reasonably anticipated. Outputs will be made available in open-access peer-reviewed publications, blogs, social media, and on a dedicated project website for future reuse.