Accenture (Italy)

The fallow deer (Dama dama) is a widely distributed cervid in Europe, often managed in enclosed settings for conservation, education, or hunting purposes. Chemical immobilization is frequently required during routine handling; however, fallow deer remain understudied in the scientific literature with regard to standardized anesthetic protocols. This study aimed to evaluate the efficacy and safety of an anesthetic protocol combining detomidine (0.2 mg/kg), ketamine (2 mg/kg), and azaperone (0.3 mg/kg) in 22 captive fallow deer. Animals were immobilized for clinical procedures, including health checks, blood sampling, individual identification, and translocation. The quality of anesthesia and physiological parameters (heart rate, respiratory rate, oxygen saturation, end-tidal CO2, arterial blood pressure, temperature) were monitored throughout immobilization. Anesthetic depth was evaluated through the assessment of reflexes, muscle tone, and eye position. The quality of induction, immobilization at approach, and recovery were systematically evaluated using dedicated scoring systems. Induction was smooth in all animals, with lateral recumbency time, and time to first safe approach achieved in 10.7 ± 6.4 min, and 22.0 ± 5.1 min, respectively. Mean handling time was 49.8 ± 9.6 min. Physiological parameters remained within acceptable limits, with only transient hypoxemia. Recovery was uneventful in most individuals, with the majority exhibiting good to excellent recovery quality and standing within 20 min of atipamezole administration. Only one individual experienced a poor-quality and prolonged recovery. No complications were observed post-procedure. The detomidine-ketamine-azaperone combination proved to be an effective and safe protocol for immobilizing fallow deer, providing stable anesthesia and good recovery quality. Dedicated scoring systems enabled a standardized and repeatable evaluation of the quality of the used protocol and its suitability. This study is the first to describe the use of azaperone in combination with detomidine and ketamine in fallow deer, contributing new insights into non-opioid immobilization strategies for cervids.

The introduction of the AI Act in the European Union presents the AI research and practice community with a set of new challenges related to compliance. While it is certain that AI practitioners will require additional guidance and tools to meet these requirements, previous research on toolkits that aim to translate the ‘theory’ of AI ethics into development and deployment ‘practice’ suggests that such resources suffer from multiple limitations. These limitations stem, in part, from the fact that the toolkits are either produced by industry-based teams (often presenting ‘ethical practice’ as a box-ticking compliance exercise) or by academics whose work tends to be abstract and divorced from the realities of industry. In this paper, we discuss the challenge of developing an AI ethics toolkit for practitioners that helps them comply with new AI-focused regulation, but that also moves beyond mere compliance to consider broader socio-ethical questions throughout development and deployment. The toolkit was created through a cross-sectoral collaboration between an academic team based in the UK and an industry team in Italy. We outline the background and rationale for creating a pro-justice AI Act compliance toolkit, detail the process undertaken to develop it, and describe the collaboration and negotiation efforts that shaped its creation. We aim for the described process to serve as a blueprint for other teams navigating the challenges of academia-industry partnerships and aspiring to produce usable and meaningful AI ethics resources.

Abstract We study how a CFO's risk‐taking incentives affect corporate hedging by utilising hand‐collected data from 2009 to 2019 on corporate hedging and managerial compensation for a sample of US oil and gas firms. The relative convexity of CFO equity compensation negatively affects the likelihood and extent of hedging. When the CFO and CEO have diverging risk‐taking incentives, the relative convexity of the CFO's equity payoff prevails over that of the CEO. This evidence underscores the primary role of the CFO in steering a firm's hedging strategy.

This investigation explores a variety of machine learning techniques, with the goal to refine the prediction of overdue payments in installment contracts associated with agricultural and commercial vehicles. The study focuses on a leading capital goods firm as a representative example. Motivated by the need for effective liquidity management, an assortment of machine learning-based models, encompassing Logistic Regression, Random Forest, XGBoost, and LightGBM, are built and evaluated using a comprehensive dataset to anticipate payment delays. The performance of each model is meticulously examined, and a feature importance investigation is carried out to pinpoint the key determinants. In a quest to boost prediction precision, ensemble methods, especially a voting classifier and an innovative merger of a neural network and a voting classifier (NN-VC), are utilized. The investigation affirms that the amalgamation of multiple machine learning algorithms significantly augments late payment forecasting, paving the way for improved risk management and insightful financial decision-making within the firm. Through the comparison of distinct models and the unveiling of a novel ensemble technique, the research offers a unique insight and sets the stage for future applications in risk management.

In this chapter we identify ten important lessons that we draw from the analysis of the Italian law and the process of its introduction and implementation, and which can also constitute a blueprint for other countries. This learning may be important for other countries that are discussing the introduction of gender quotas, for policy-makers, for corporations and their executive directors, who will have to include women in the top management and on the boards of their companies, for academics and researchers, experts in corporate governance and for women’s organizations. Before moving to our ‘decalogue’, this is a brief summary of the analysis developed in the book.

The mixed-potential integral-equation approach, using spatial-domain closed-form Green's functions, and discretized with the method-of-moments, is a state-of-the-art method for the analysis of planar microstrip circuits. One of its most time-demanding tasks is the evaluation of the impedance matrix terms, which typically requires the numerical computation of two-dimensional integrals. A method based on suitable changes of coordinates and domains is introduced in this paper in order to reduce such integrals to a quasi-one-dimensional numerical integration, with a substantial enhancement in the efficiency of the analysis, without affecting the accuracy of the approach. Results are given demonstrating, for practical accuracy values, an improvement of typically one order of magnitude in simulation times.

This paper proposes a new vector quantization based (VQ-based) technique for very low bit rate encoding of multispectral images. We rely on the assumption that the shape of a generic spatial block does not change significantly from band to band, as is the case for high spectral-resolution imagery. In such a hypothesis, it is possible to accurately quantize a three-dimensional (3-D) block-composed of homologous two-dimensional (2-D) blocks drawn from several bands-as the Kronecker-product of a spatial-shape codevector and a spectral-gain codevector, with significant computation saving with respect to straight VQ. An even higher complexity reduction is obtained by representing each 3-D block in its minimum-square-error Kronecker-product form and by quantizing the component shape and gain vectors. For the block sizes considered, this encoding strategy is over 100 times more computationally efficient than unconstrained VQ, and over ten times more computationally efficient than direct gain-shape VQ. The proposed technique is obviously suboptimal with respect to VQ, but the huge complexity reduction allows one to use much larger blocks than usual and to better exploit both the statistical and psychovisual redundancy of the image. Numerical experiments show fully satisfactory results whenever the shape-invariance hypothesis turns out to be accurate enough, as in the case of hyperspectral images. In particular, for a given level of complexity and image quality, the compression ratio is up to five times larger than that provided by ordinary VQ, and also larger than that provided by other techniques specifically designed for multispectral image coding.