Flock (United Kingdom)

Beef and sheep enterprises can present challenges for data capture and analysis, which can be demotivating for both farmers and veterinarians. Veterinarians are well placed to address these challenges through skills in data analysis and knowledge of the farm system, enterprise aims and farmer preferences. ‘Easy options' can be used as a starting point, such as counting wet, dry and field dead lambs with a bucket system or using movement data to assess calving patterns. technicianniciannology exists to allow data capture and analysis, with a variety of options available from basic to more advanced, depending on requirements. Antimicrobial sales data are also available to veterinarians. These data can be entered into the AHDB Medicine Hub to calculate antibiotic-use metrics for use at both the individual farm and ruminant-sector level.

Federated learning (FL) is a machine learning paradigm, which enables multiple and decentralized clients to collaboratively train a model under the orchestration of a central aggregator. FL can be a scalable machine learning solution in <italic>big data</italic> scenarios. Traditional FL relies on the trust assumption of the central aggregator, which forms cohorts of clients honestly. However, a malicious aggregator, in reality, could abandon and replace the client's training models, or insert fake clients, to manipulate the final training results. In this work, we introduce <monospace>zkFL</monospace>, which leverages zero-knowledge proofs to tackle the issue of a malicious aggregator during the training model aggregation process. To guarantee the correct aggregation results, the aggregator provides a proof per round, demonstrating to the clients that the aggregator executes the intended behavior faithfully. To further reduce the verification cost of clients, we use blockchain to handle the proof in a zero-knowledge way, where miners (<italic>i.e.</italic>, the participants validating and maintaining the blockchain data) can verify the proof without knowing the clients' local and aggregated models. The theoretical analysis and empirical results show that <monospace>zkFL</monospace> achieves better security and privacy than traditional FL, without modifying the underlying FL network structure or heavily compromising the training speed. Authors

When working with AI applications and models, the privacy of individuals is of utmost importance. Privacy in machine learning aims to preserve the privacy of individuals represented in the datasets. It tends to employ and use different techniques to safeguard sensitive data during ML development and deployment. In this chapter, we’ll look at privacy and how to measure privacy in ML models and algorithms.

In the preceding chapter, we looked at an important and not very popular responsible AI component, sustainability, using the ESG factors. In this chapter, we’ll discuss another important topic in responsible AI, known as human-centered AI or HCAI.

In the previous chapter, we looked at AI governance, a few AI regulations and policies such as the EU AI Act, Joe Biden’s US executive order, and a couple of reports from the UN and UNESCO. We also introduced the SAFE-HAI framework which is critical to achieving responsibly built AI algorithms that adheres to the newly introduced AI policies. In this chapter, we’ll review “accuracy,” a key part of the SAFE-HAI framework and a crucial responsible AI principle. We’ll briefly discuss its definition, key questions to consider regarding accuracy and performance, ways to measure accuracy, a scoring rubric for accuracy, a few benchmarks from the research community, and ways to resolve poor accuracy of ML models and applications.

BackgroundManagement of pain is critical to improve the welfare of farmed livestock and meet consumer expectations. There is limited published information about the use of analgesic drugs in the sheep sector.MethodsA mixed‐method approach was followed. The range of analgesic drugs used on 52 Northern Irish sheep farms was determined through analysis of medicine purchase records. Through interview and discussion groups, with both farmer and veterinarian participants, attitudes towards the use and adoption of such medicines were explored.ResultsThe use of non‐steroidal anti‐inflammatory drugs (NSAIDs) was widespread and highly variable. One‐third of farmers in the sample did not purchase any NSAID. Meloxicam was the most commonly purchased NSAID by mass (72%) and standardised dose (73%). During interviews and discussions, farmers outlined the benefits they saw in using NSAIDs and how veterinarians influenced their uptake of these medicines. Use of corticosteroid was evidenced on 50% of the farms that supplied medicine records for analysis.ConclusionsVeterinarians can influence farmers to adopt NSAIDs for the provision of analgesia in their sheep and farmers observed the benefits they delivered. However, many farmers are still to be reached with this message, perhaps due to being largely self‐sufficient and rarely engaging with veterinarians. Veterinarians have the opportunity to challenge farmers about the provision of analgesia, especially when farmers seek antibiotics for painful conditions such as lameness. Currently, the lack of an authorised product in the UK, with associated treatment guidance and industry promotion, may limit veterinarians’ confidence in prescribing drugs for pain control in sheep.

Dogs (Canis lupus familiaris) are one of the most common pets around the world but ownership patterns and human-dog interactions have been changing, particularly in developing nations. We conducted household surveys in Costa Rica to characterize dog ownership, the owned dog population, where dogs were confined at night and in the morning, and behaviors regarding selected dog care issues. We also compared these results to similar questionnaires used in Costa Rica over the past 20 years. We found 76% of households in Costa Rica owned at least one dog and on average there were about 1.4 dogs owned per household. These dog ownership rates are higher than previous estimates. The probability of owning a dog was highest on farms and lowest in single family dwellings without a yard, higher among respondents that owned their homes and decreasing with increasing human population density The total number of owned dogs in Costa Rica was estimated to be 2,222,032 (95% confidence intervals: 1,981,497–2,503,751). The sterilization rate for homed dogs in 2020 was approximately 62% (females: 67%, males: 61%) which is higher than the 18% of owned dogs that were sterilized in a 2003 survey. Overall, only 1.2% (95% CI: 0.3–2.5%) of owned dogs slept on the street with a slightly higher proportion on the street at 8 am. The number of owned dogs roaming the streets at night nation-wide was estimated to be 27,208 (95% CI: 7,557–56,619) compared to 43,142 (95% CI: 20,118–73,618) on the street at 8 am. The number of unowned free-roaming dogs in Costa Rica has never been estimated but we can generate some idea of the size of the unowned dog population by determining the proportion of free-roaming dogs on the street wearing collars. There was a negative relationship between human population density and owned dogs being on the street meaning fewer dogs roam the streets in highly populated areas compared to less populated areas. Overall, we identify trends against which future progress can be measured and provide information that are critical in designing effective humane dog management programs in Costa Rica in the future.

Sheep scab (psoroptic mange), which is endemic in the United Kingdom (UK) flock, has a significant, negative impact on sheep welfare. Nothing has previously been published about the distribution of sheep scab in Northern Ireland (NI), nor about Northern Irish farmers’ knowledge and behaviours relating to the disease, its treatment, prevention and control. Between March and June 2021 an online questionnaire on the disease was completed by sheep farmers in NI. Forty-four respondents out of a total of 122 valid returns (36%) indicated that they had at least one outbreak of sheep scab in their flock within the previous five years. These flocks were spread throughout NI and included flocks grazing on common land. Farmers reporting sheep scab in their flock considered movements of sheep between flocks to be the main cause of flock infestation. Respondents demonstrated knowledge gaps in relation to the parasite biology, disease transmission, prevention and treatment options, as well as a lack of awareness of some of the relevant industry guidelines. We highlight that some farmers rely on clinical signs alone to rule out the possibility that newly purchased sheep are infested with sheep scab before mixing them with their flock. This activity poses a high risk for the introduction of sheep scab into previously uninfested flocks. The inadequacy of some farmers’ quarantine rules, or their inability to follow them, was also reported by farmers as being the cause of their flock infestation. Sheep scab outbreaks were shown to result in significant financial cost, with some farmers reporting their most recent outbreak had cost over £2500 ($3329). The paper also highlights that in addition to the animal health and welfare impact and financial cost, sheep scab was reported to have a social cost: 94 respondents (79%) agreed that a sheep scab outbreak caused emotional stress to affected farmers. These findings have provided evidence of the widespread nature of sheep scab in the NI flock, and of the knowledge gaps and behaviours which need to be addressed to improve sheep scab control. This will require a combination of focused research, knowledge exchange between farmers, advisors, policy makers and regulators, and co-developed disease control plans at a flock and national level.