Operator and organizational maintenance manual : pumping assembly, flammable liquid, bulk transfer, lightweight, centrifugal, 100 gpm minimum, gasoline-engine-driven (Barnes model US6ACG), FSN 4320-150-6116.

Includes index.

Operator, organizational, direct, and general support maintenance manual : pump assembly, flammable liquid, bulk transfer, gasoline engine driven, 350-gpm capacity at 190 foot head, wheel mounted (Gorman-Rupp model 84C15-4A084) FSN 4230-916-9172 and (Barnes model US36ACG) FSN 4320-407-2583.

Includes index.

Operation and organizational maintenance manual : pumping assembly, flammable liquid, bulk transfer, gasoline engine driven, 350 gpm capacity, 275 feet total dynamic head, wheel mounted (Peabody Barnes, Inc. model US37ACG), FSN 4320-195-4914.

Includes index.

Operator, organizational, direct and general suport, and depot maintenance manual : pump, centrifugal, gasoline engine driven, 50 GPM, 100-foot head, flammable liquid, bulk transfer (Barnes model-QM-2-28002) FSN 4320-913-7131.

Includes index.

Organizational, direct support, and general support maintenance repair parts and special tools list (including depot maintenance repair parts and special tools) : pumping assembly, flammable liquid, bulk transfer : gasoline engine driven; 350 GPM capacity at 190 feethead; wheel mounted (Gorman-Rupp model 84C15-4A084) NSN 4320-00-916-9172 and (Barnes model US36ACG) NSN 4320-00-407-2583.

Includes index.

Operator, organizational, direct and general support : pumping assembly, flammable liquid, bulk transfer, 50 GPM, (Barnes model 9117CA) Army model SPE 16 FSN 4320-728-0265, (Barnes model 15671CA)[ Army model SPE 16A FSN 4320-658-2888, (Barnes model 4074CA) FSN 4320-271-1858.

"May 1968."

Operation and organizational maintenance manual : pumping assembly, flammable liquid, bulk transfer, gasoline engine driven, 350 gpm capacity, 275 feet total dynamic head, wheel mounted (Peabody Barnes, Inc. model US37ACG), FSN 4320-195-4914.

"January 1974."

The traditional rationale of the arm's length approach to transfer pricing - should the separate accounting model be maintained for modern multinational entities?

This article examines the current transfer pricing regime to consider whether it is a sound model to be applied to modern multinational entities. The arm's length price methodology is examined to enable a discussion of the arguments in favour of such a regime. The article then refutes these arguments concluding that, contrary to the very reason multinational entities exist, applying arm's length rules involves a legal fiction of imagining transactions between unrelated parties. Multinational entities exist to operate in a way that independent entities would not, which the arm's length rules fail to take into account. As such, there is clearly an air of artificiality in applying the arm's length standard. To demonstrate this artificiality with respect to modern multinational entities, multinational banks are used as an example. The article concluded that the separate entity paradigm adopted by the traditional transfer pricing regime is incongruous with the economic theory of modern multinational enterprises.

Inter-organisational knowledge transfer process model

Knowledge management (KM) is an emerging discipline (Ives, Torrey & Gordon, 1997) and characterised by four processes: generation, codification, transfer, and application (Alavi & Leidner, 2001). Completing the loop, knowledge transfer is regarded as a precursor to knowledge creation (Nonaka & Takeuchi, 1995) and thus forms an essential part of the knowledge management process. The understanding of how knowledge is transferred is very important for explaining the evolution and change in institutions, organisations, technology, and economy. However, knowledge transfer is often found to be laborious, time consuming, complicated, and difficult to understand (Huber, 2001; Szulanski, 2000). It has received negligible systematic attention (Huber, 2001; Szulanski, 2000), thus we know little about it (Huber, 2001). However, some literature, such as Davenport and Prusak (1998) and Shariq (1999), has attempted to address knowledge transfer within an organisation, but studies on inter-organisational knowledge transfer are still much neglected. An emergent view is that it may be beneficial for organisations if more research can be done to help them understand and, thus, to improve their inter-organisational knowledge transfer process. Therefore, this article aims to provide an overview of the inter-organisational knowledge transfer and its related literature and present a proposed inter-organisational knowledge transfer process model based on theoretical and empirical studies.

Reduced-parameter model of head-related transfer functions for synthesized spatial audio

Digital systems can generate left and right audio channels that create the effect of virtual sound source placement (spatialization) by processing an audio signal through pairs of Head-Related Transfer Functions (HRTFs) or, equivalently, Head-Related Impulse Responses (HRIRs). The spatialization effect is better when individually-measured HRTFs or HRIRs are used than when generic ones (e.g., from a mannequin) are used. However, the measurement process is not available to the majority of users. There is ongoing interest to find mechanisms to customize HRTFs or HRIRs to a specific user, in order to achieve an improved spatialization effect for that subject. Unfortunately, the current models used for HRTFs and HRIRs contain over a hundred parameters and none of those parameters can be easily related to the characteristics of the subject. This dissertation proposes an alternative model for the representation of HRTFs, which contains at most 30 parameters, all of which have a defined functional significance. It also presents methods to obtain the value of parameters in the model to make it approximately equivalent to an individually-measured HRTF. This conversion is achieved by the systematic deconstruction of HRIR sequences through an augmented version of the Hankel Total Least Squares (HTLS) decomposition approach. An average 95% match (fit) was observed between the original HRIRs and those re-constructed from the Damped and Delayed Sinusoids (DDSs) found by the decomposition process, for ipsilateral source locations. The dissertation also introduces and evaluates an HRIR customization procedure, based on a multilinear model implemented through a 3-mode tensor, for mapping of anatomical data from the subjects to the HRIR sequences at different sound source locations. This model uses the Higher-Order Singular Value Decomposition (HOSVD) method to represent the HRIRs and is capable of generating customized HRIRs from easily attainable anatomical measurements of a new intended user of the system. Listening tests were performed to compare the spatialization performance of customized, generic and individually-measured HRIRs when they are used for synthesized spatial audio. Statistical analysis of the results confirms that the type of HRIRs used for spatialization is a significant factor in the spatialization success, with the customized HRIRs yielding better results than generic HRIRs.

Development of a porcine lactation model for the evaluation of mammary transfer of exogenous molecules - A contribution from the conception project

Around 5 million women give birth each year in Europe and, while breastfeeding, the majority of them may need to take medications, either occasionally or continuously. Unfortunately, there is often scarce evidence of trustworthy information about how a specific molecule might affect the physiology of lactation. This is the reason that brought a European public-private partnership to fund the development of a reliable platform to provide women and health-care professionals a helpful instrument to reduce uncertainty about the effects of medication used during breastfeeding. On April 1st 2019, the ConcePTION project (Grant Agreement n°821520) started to develop such envisaged platform. The 3rd Work Package was in charge of the validation of in vitro, in vivo and in silico lactation models. Between the numerous species currently used in preclinical studies, pigs’ similarities with humans’ anatomy, physiology and genomics make them extremely useful as translational models, when proper veterinary expertise is applied. The ASA team from the University of Bologna, went first to characterize the translational lactation model using the swine species, chosen upon literature review. The aim of this work was to lay the foundations of a porcine lactation model that could be suitable for application within pharmaceutical tests, to study drug transfer through milk prior approval and commercialization. The obtained results highlighted both strengths and critical points of the study design, allowing a significant improvement in the knowledge of pharmacokinetic physiology in lactating mammals. Lastly, this project allowed the assessment of microbial changes in gut resident bacteria of newborns through an innovative in vitro colonic model. Indeed, even if there were no evident adverse effects determined by drug residues in milk, possible alterations in the delicate microbial ecology of newborns’ gastrointestinal tract was considered pivotal, giving its possible impact on the individual health and growth.

Vocabulary Transfer and Knowledge Distillation for Language Model Compression

Natural Language Processing (NLP) has seen tremendous improvements over the last few years. Transformer architectures achieved impressive results in almost any NLP task, such as Text Classification, Machine Translation, and Language Generation. As time went by, transformers continued to improve thanks to larger corpora and bigger networks, reaching hundreds of billions of parameters. Training and deploying such large models has become prohibitively expensive, such that only big high tech companies can afford to train those models. Therefore, a lot of research has been dedicated to reducing a model’s size. In this thesis, we investigate the effects of Vocabulary Transfer and Knowledge Distillation for compressing large Language Models. The goal is to combine these two methodologies to further compress models without significant loss of performance. In particular, we designed different combination strategies and conducted a series of experiments on different vertical domains (medical, legal, news) and downstream tasks (Text Classification and Named Entity Recognition). Four different methods involving Vocabulary Transfer (VIPI) with and without a Masked Language Modelling (MLM) step and with and without Knowledge Distillation are compared against a baseline that assigns random vectors to new elements of the vocabulary. Results indicate that VIPI effectively transfers information of the original vocabulary and that MLM is beneficial. It is also noted that both vocabulary transfer and knowledge distillation are orthogonal to one another and may be applied jointly. The application of knowledge distillation first before subsequently applying vocabulary transfer is recommended. Finally, model performance due to vocabulary transfer does not always show a consistent trend as the vocabulary size is reduced. Hence, the choice of vocabulary size should be empirically selected by evaluation on the downstream task similar to hyperparameter tuning.