9,845 research outputs found
Metadata Representations for Queryable ML Model Zoos
No full textMachine learning (ML) practitioners and organizations are building model zoos of pre-trained models, containing metadata describing properties of the ML models and datasets that are useful for reporting, auditing, reproducibility, and interpretability purposes. The metatada is currently not standardised; its expressivity is limited; and there is no interoperable way to store and query it. Consequently, model search, reuse, comparison, and composition are hindered. In this paper, we advocate for standardized ML model metadata representation and management, proposing a toolkit supported to help practitioners manage and query that metadata.Web Information SystemsHuman-Centred Artificial Intelligenc
A Manifesto of Nodalism
No full textThis paper proposes the notion of Nodalism as a means describing contemporary culture and of understanding my own creative practice in electronic music composition. It draws on theories and ideas from Kirby, Bauman, Bourriaud, Deleuze, Guatarri, and Gochenour, to demonstrate how networks of ideas or connectionist neural models of cognitive behaviour can be used to contextualize, understand and become a creative tool for the creation of contemporary electronic music
Optimizing ML Inference Queries Under Constraints
No full textThe proliferation of pre-trained ML models in public Web-based model zoos facilitates the engineering of ML pipelines to address complex inference queries over datasets and streams of unstructured content. Constructing optimal plan for a query is hard, especially when constraints (e.g. accuracy or execution time) must be taken into consideration, and the complexity of the inference query increases. To address this issue, we propose a method for optimizing ML inference queries that selects the most suitable ML models to use, as well as the order in which those models are executed. We formally define the constraint-based ML inference query optimization problem, formulate it as a Mixed Integer Programming (MIP) problem, and develop an optimizer that maximizes accuracy given constraints. This optimizer is capable of navigating a large search space to identify optimal query plans on various model zoos.Green Open Access added to TU Delft Institutional Repository ‘You share, we take care!’ – Taverne project https://www.openaccess.nl/en/you-share-we-take-care Otherwise as indicated in the copyright section: the publisher is the copyright holder of this work and the author uses the Dutch legislation to make this work public.Web Information SystemsHuman-Centred Artificial Intelligenc
Primary biliary cirrhosis is associated to the HLA DRB1*0301-DQA1*0501-DQB1*0201 haplotype in Sardinians
No full textPrimary biliary cirrhosis is associated to the HLA DRB1*0301-DQA1*0501-DQB1*0201 haplotype in Sardinians
No full text
Impact of the direct-acting antiviral agents (DAAs) on chronic hepatitis C in Sardinian patients with transfusion-dependent Thalassemia major
No full textBackground and aims: Direct antiviral agents (DAAs) have revolutionised the standard of care for the treatment of hepatitis even in patients with hemoglobinopathies. The aim of this study is to show how, thanks to DAAs, HCV infection has been substantially eradicated in one of the biggest Centres for the management of Thalassemia in Europe. Methods: Thalassemia major patients regularly transfused and iron chelated in Cagliari (Italy) who were HCV-RNA positive were evaluated for the potential prescription of antiviral therapy. Results: A total of 99 patients, 26 of whom had been diagnosed with cirrhosis, were treated with at least one dose of DAAs, which proved to be safe and well tolerated. Two of the patients died during the treatment after becoming HCV-RNA negative while another voluntarily interrupted the therapy. The final SVR in the patients who completed the treatment was 100%, while measuring 97% (96/99) in the Intention-to-Treat analysis. After DAAs, no new cases of hepatocellular carcinoma have been reported. Conclusions: The use of DAAs in patients suffering from beta-Thalassemia major with chronic hepatitis C or cirrhosis can be considered safe and effective. Close monitoring for hepatocellular carcinoma development is, in any case, recommended indefinitely post-SVR
Antipsychotics and Torsadogenic Risk: Signals Emerging from the US FDA Adverse Event Reporting System Database
No full textBackground: Drug-induced torsades de pointes (TdP) and related clinical entities represent a current regulatory and clinical burden. Objective: As part of the FP7 ARITMO (Arrhythmogenic Potential of Drugs) project, we explored the publicly available US FDA Adverse Event Reporting System (FAERS) database to detect signals of torsadogenicity for antipsychotics (APs). Methods: Four groups of events in decreasing order of drug-attributable risk were identified: (1) TdP, (2) QT-interval abnormalities, (3) ventricular fibrillation/tachycardia, and (4) sudden cardiac death. The reporting odds ratio (ROR) with 95 % confidence interval (CI) was calculated through a cumulative analysis from group 1 to 4. For groups 1+2, ROR was adjusted for age, gender, and concomitant drugs (e.g., antiarrhythmics) and stratified for AZCERT drugs, lists I and II (http://www.azcert.org, as of June 2011). A potential signal of torsadogenicity was defined if a drug met all the following criteria: (a) four or more cases in group 1+2; (b) significant ROR in group 1+2 that persists through the cumulative approach; (c) significant adjusted ROR for group 1+2 in the stratum without AZCERT drugs; (d) not included in AZCERT lists (as of June 2011). Results: Over the 7-year period, 37 APs were reported in 4,794 cases of arrhythmia: 140 (group 1), 883 (group 2), 1,651 (group 3), and 2,120 (group 4). Based on our criteria, the following potential signals of torsadogenicity were found: amisulpride (25 cases; adjusted ROR in the stratum without AZCERT drugs = 43.94, 95 % CI 22.82-84.60), cyamemazine (11; 15.48, 6.87-34.91), and olanzapine (189; 7.74, 6.45-9.30). Conclusions: This pharmacovigilance analysis on the FAERS found 3 potential signals of torsadogenicity for drugs previously unknown for this risk
Building a generalisable ML pipeline at ING
No full textAdvances in data science have caused an increase in the use of Artificial Intelligence (AI), specifically Machine Learning (ML), throughout various fields. Not only in research but in the industry as well, has ML been receiving increasing amounts of interest. Many companies rely on ML models to increase the efficiency of existing processes or offer new services and products. The industry, however, is facing several additional challenges compared to the academic context. One of those challenges is applying the Development Operations (DevOps) model to an ML application, also referred to as MLOps. This thesis sets out to find the specific challenges that practitioners encounter while operationalising ML models. To do so, we perform a single-case case study on an ML pipeline built by the Trade & Communication Surveillance team at the ING bank. This case study consists of conducting a set of interviews and performing a manual code inspection of the pipeline. The team faces challenges ranging from having insufficient time for operationalising each ML project individually to operating in the highlyregulated fintech context. Their pipeline is able to deploy a single ML model but it does not generalise well to other projects. We present the first version of an application that mitigates these challenges. The application is able to deploy ML models to the development environment at ING and can be operated by data scientists to reduce the effort of operationalising an ML model. Computer Science | Software Technolog
- …
