9,715 research outputs found
Efficienza ed economicità di un device centrale a inserzione periferica
No full textIn questo studio l’attenzione è focalizzata sul passaggio culturale da una logica di re-active a un adeguato sistema di pro-active,
dove uno dei parametri da valutare nel comportamento di un infermiere è quello relativo alla sua capacità di formulare una diagnosi precoce
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
Assessing the Stock Dynamics of Elasmobranchii off the Southern Coast of Sicily by Using Trawl Survey Data
No full textElasmobranchii (sharks and rays), which have peculiar and vulnerable life-history traits, are highly threatened by fishing activities. Indeed, between 53% and 71% of Mediterranean elasmobranch species are at risk of extinction. In this context, using the abundance MSY (AMSY) model, the present study provides an assessment of 20 batoids and 16 shark species in the Strait of Sicily, sampled during a bottom trawl survey from 1995 to 2020. Overall, the outputs underline a progressively improving condition for shark and ray assemblages of both shelf and eurybathic zones. As for slope-dwelling species, a horseshoe-shaped dynamic, characterized by a progressive decrease in relative harvesting pressure and an increase in relative biomass followed by an increase in fishing pressure and decrease in biomass, was detected. The dynamics of the Elasmobranchii living in the Strait of Sicily appear to be affected by changes in the fishing patterns of trawlers, showing a shift from shallow water to bathyal fishing grounds and targeting deep-water red shrimp. In this context, it seems wise to limit the impact of deep-water fisheries on Elasmobranchii by reducing fishing efforts and implementing ad hoc management measures aimed at safeguarding these vulnerable species
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
Guarding net effects on landings and discards in Mediterranean trammel net fishery: Case analysis of Egadi Islands Marine Protected Area (Central Mediterranean Sea, Italy)
No full textDiscards remain among the main negative impacts of fishing activities, and their reductions are strengthened by the European Common Fisheries Policy (European Regulation 1380/2013). Trammel net fisheries appear more sustainable compared with other fishing techniques, especially from an ecological viewpoint. Despite this, reports show that trammel net fisheries deliver discard quantities between 10% and 43% of the total catch biomass. To supplement existing information, this current work attempts to address the discard reduction using guarding net in the small-scale fisheries of Egadi Islands MPA (Western Sicily, Central Mediterranean Sea). To assess the reduction of unwanted catches, 48 experimental fishing trials were conducted within a 6-month period. The experimental fishing trial employed a trammel net made up of 20 panels alternated with two different net configurations. The control panels (CN) held a large outer (180 mm) and small inner (31.25 mm) meshes. The test panels (GN) with guarding net constituted a three-mesh-high (50-mm mesh size) net placed between trammel net panels and a lead line. A total of 3,310 individuals belonging to 106 taxa and nine phyla were caught. Crustaceans were the most abundant unwanted catches in the control panels, whereas bioconstructions occurred in the guarding net panels. The discard ratios of CN and GN panels were statistically different (t-value = -2.55; p< 0.05). The analysis of catch per unit effort showed higher catches of CN panels for both commercial and discard fractions (p< 0.05). Moreover, the guarding net panels caught the main discarded species at 20% lower compared with the control. The overall value of the catch at the CN panels (euro 3,366.90) was higher than the total income (euro 2,043.70) generated using the GN panels, which suggests a significant commercial loss of 40% (p< 0.05)
Evaluation of the Economic Performance of Coastal Trawling off the Southern Coast of Sicily (Central Mediterranean Sea)
No full textThe economic performances of four trawling fleets (those of the Sicilian cities of Trapani, Sciacca, Licata and Porto Palo di Capo Passero) operating in the coastal waters along the southern coast of Sicily (geographical Subarea 16), and potentially affected by the establishment of the Fisheries Restricted Areas (FRAs), were analysed. The main economic performance results (revenues, costs and profits) of 37 trawlers were calculated prior to the implementation of FRAs and compared with those estimated by the spatial bio-economic model SMART after the FRAs’ establishment. Results showed that the fleets of Sciacca and Licata, located in the central part of the southern Sicilian coast, had a short-term reduction of profits as a result of the implementation of the FRAs; conversely, a short-term increase in the economic performances of Trapani and Porto Palo di Capo Passero fleets was expected. Although the FRAs represent a good tool for rebuilding overexploited stocks, the different socio-economic impacts of the single fleets should be assessed before adopting them and the implementation of specific compensative measures should be planned for the impacted fleet until a more productive state of the stock is reached
First record and otolith morphometric description of an adult lightfish, Ichthyococcus ovatus (Actinopterygii: Stomiiformes: Phosichthyidae), caught in the Strait of Sicily (central Mediterranean Sea)
No full textOn July 2018, one specimen ofIchthyococcus ovatus (Cocco, 1838) was caught in the Strait of Sicily during the International Bottom Trawl Survey in the Mediterranean (MEDITS). The adult I. ovatus measured 49 mm in total length and weighed 1.44 g. In this context, the presently reported study constitutes the first and deepest record of an adult of I. ovatus as well as the morphometric description of its sagittal otoliths. In addition, we provide an age estimation as well as an update of the geographical distribution of this bathypelagic species around the Mediterranean Sea. Based on the growth increments of sagittal otoliths, the estimated age was five years. Specifically, the otolith from the presently reported specimen of I. ovatus tended to be elliptic in shape related to aspect ratio and high rectangularity while circularity showed high complexity of otolith contour complexity. The absence of economic val-ue of rarely reported species may underestimate their abundance. Therefore, more studies and research surveys would be necessary to fill the information gap on the biology of these deep-water species
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
'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
- …
