9,755 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
Estimation of insulin sensitivity and glucose clearance from minimal model: new insights from labelled IVGTT.
No full textThe "minimal model" of glucose disappearance provides noninvasive estimates of insulin sensitivity and glucose effectiveness from an intravenous glucose tolerance test (IVGTT). However, this model does not allow the separation of glucose production from utilization. To overcome this limitation, labeled glucose was injected along with cold glucose in six normal dogs, and both cold and labeled glucose time courses were monitored along with insulin concentration. A revised minimal model was fitted to tracer data to obtain new measures of insulin sensitivity (SI* = 6.41 +/- 0.91 10(-4) min-1 X microU-1 X ml-1) and fractional glucose clearance (SG* = 0.0092 +/- 0.0009 min-1). SG* was compared with a direct measure obtained by a hepatic arterial-venous difference technique, which yielded a value of 0.0097 +/- 0.0002, virtually identical to SG*, thereby validating the model estimate. When the original minimal model was identified from cold data, we obtained S1 = 4.52 +/- 1.39 and SG = 0.042 +/- 0.009. SI* and SG* were different from SI and SG, respectively. In particular SG overestimates fractional glucose clearance by approximately five times. The revised minimal model yields glucose disposal parameters SI* and SG* that are not affected by the confounding effect of insulin and glucose inhibition of glucose production. Limitations inherent in cold IVGTT and original minimal model are overcome by labeled IVGTT and the revised minimal model, while test simplicity remains
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
Vascular smooth muscle cell dysfunction in patients with migraine.
No full textBackground: Migraine is associated with increased risk of cardiovascular disease, but the mechanisms are unclear.Objective: To investigate the activity of endothelial and vascular smooth muscle cells (VSMCs) in patients with migraine.Methods: Case-control study of 12 patients with migraine without aura and 12 matched healthy control subjects. Endothelial and VSMC components of vascular reactivity were explored by plethysmography measurement of forearm blood flow (FBF) during infusions of vasoactive agents into the brachial artery. Forearm production of nitric oxide (NO) and cyclic guanosine monophosphate (cGMP) was also quantified.Results: In patients with migraine, the vasodilating effect of acetylcholine (ACh), an endothelium-ependent vasodilator, was markedly reduced (p < 0.001 by analysis of variance). In response to the highest dose of ACh, FBF rose to 8.6 +/- 2.2 in patients with migraine and to 22.7 +/- 3.0 mL x dL(-1) x min(-1) in controls (p = 0.001). The dose-response curve to nitroprusside, a vasodilator directly acting on VSMCs, was depressed in patients with migraine (p < 0.001 by analysis of variance). The maximal response of FBF to nitroprusside was 12.1 +/- 2.0 in patients with migraine and 24.1 +/- 1.8 mL x dl(-1) x min(-1) in controls (p < 0.001). During ACh infusion, NO release from the endothelium was similar in patients and controls. In contrast, there was a marked release of cGMP from VSMCs in controls, but not in patients with migraine (-1.9 +/- 2.2 in patients with migraine and -19.1 +/- 5.4 nmol x dL(-1) x min(-1) in controls; p = 0.03).Conclusions: Patients with migraine are characterized by a distinct vascular smooth muscle cell dysfunction, revealed by impaired cyclic guanosine monophosphate and hemodynamic response to nitric oxide
'Project smells' - Experiences in Analysing the Software Quality of ML Projects with mllint
No full textMachine Learning (ML) projects incur novel challenges in their development and productionisation over traditional software applications, though established principles and best practices in ensuring the project's software quality still apply. While using static analysis to catch code smells has been shown to improve software quality attributes, it is only a small piece of the software quality puzzle, especially in the case of ML projects given their additional challenges and lower degree of Software Engineering (SE) experience in the data scientists that develop them. We introduce the novel concept of project smells which consider deficits in project management as a more holistic perspective on software quality in ML projects. An open-source static analysis tool mllint was also implemented to help detect and mitigate these. Our research evaluates this novel concept of project smells in the industrial context of ING, a global bank and large software- and data-intensive organisation. We also investigate the perceived importance of these project smells for proof-of-concept versus production-ready ML projects, as well as the perceived obstructions and benefits to using static analysis tools such as mllint. Our findings indicate a need for context-aware static analysis tools, that fit the needs of the project at its current stage of development, while requiring minimal configuration effort from the user. 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.Software EngineeringSoftware Technolog
Audiomobiles, Sculptures and Conundrums
No full textRoberto Gerhard was a pioneer of electronic music in England creating a number of substantial concert, theatre and radio works from as early as 1954. Gerhard’s electronic music is one of the richest repositories for understanding the development of the composer’s late compositional technique. Apart from the Symphony no.3, ‘Collages’, none of Gerhard’s electronic music is published. This paper will discuss aspects of Gerhard’s electronic music, focusing on Audiomobiles (1958-59) and Sculptures (1963)
Differential roles of splanchnic and peripheral tissues in the pathogenesis ofimpaired glucose tolerance
No full textTo identify the mechanism(s) of the altered glucoregulatory response to a glucose
load in subjects with impaired glucose tolerance, we selectively quantitated the
components of net splanchnic glucose balance, i.e., splanchnic glucose uptake and
hepatic glucose output, as well as peripheral glucose uptake, by combining
[3-3H]glucose infusion with hepatic vein catheterization. After intravenous
glucose infusion (6 mg X kg-1 X min-1 for 90 min), blood glucose rose to 172 +/-
7 mg/dl in controls and 232 +/- 13 mg/dl in subjects with impaired glucose
tolerance (P less than 0.01). The response of plasma insulin did not differ
significantly between the two groups (29 +/- 4 vs. 40 +/- 10 microU/ml at 90 min
in control and in glucose intolerant subjects, respectively; P = NS). In both
groups, glucose infusion caused the net splanchnic glucose balance to switch from
the net output of the basal state to a net glucose uptake. However, this effect
was more marked in subjects with impaired glucose tolerance than in control
subjects (at 90 min: 2.83 +/- 0.53 vs. 1.60 +/- 0.18 mg X kg-1 X min-1,
respectively: P less than 0.05). The different pattern of splanchnic glucose
balance was entirely accounted for by a greater rise in splanchnic glucose uptake
in the group of glucose intolerants , as the suppression of endogenous glucose
output by the glucose load was practically complete in both groups. In contrast,
glucose uptake by peripheral tissues increased considerably less in subjects with
impaired glucose tolerance than in controls (2.2-2.6 vs 3.6-4.1 mg X kg-1 X
min-1, respectively, between 60 and 90 min; P less than 0.01-0.001). Furthermore,
a net splanchnic lactate uptake was present in the basal state, which was
inhibited by the glucose load and switched to a comparable net lactate output in
both groups. These results indicate that the mechanism responsible for the
altered glucoregulation in subjects with impaired glucose tolerance resides
entirely in the peripheral tissues whose ability to dispose of a glucose load is
drastically reduced. On the other hand, no defect is detectable in any of the
explored mechanisms regulating splanchnic glucose metabolism during the disposal
of an exogenous glucose load
- …
