177 resultados para improved SVM
Resumo:
Immunomagnetic separation (IMS) represents a simple but effective method of selectively capturing and concentrating Mycobacterium bovis, the causative agent of bovine tuberculosis (bTB), from tissue samples. It is a physical cell separation technique that does not impact cell viability, unlike traditional chemical decontamination prior to culture. IMS is performed with paramagnetic beads coated with M. bovis-specific antibody and peptide binders. Once captured by IMS, M. bovis cells can be detected by either PCR or cultural detection methods. Increased detection rates of M. bovis, particularly from non-visibly lesioned lymph node tissues from bTB reactor animals, have recently been reported when IMS-based methods were employed.
Resumo:
In modern semiconductor manufacturing facilities maintenance strategies are increasingly shifting from traditional preventive maintenance (PM) based approaches to more efficient and sustainable predictive maintenance (PdM) approaches. This paper describes the development of such an online PdM module for the endpoint detection system of an ion beam etch tool in semiconductor manufacturing. The developed system uses optical emission spectroscopy (OES) data from the endpoint detection system to estimate the RUL of lenses, a key detector component that degrades over time. Simulation studies for historical data for the use case demonstrate the effectiveness of the proposed PdM solution and the potential for improved sustainability that it affords.
Resumo:
Unlike the mathematical encryption and decryption adopted in the classical cryptographic technology at the higher protocol layers, it is shown that characteristics intrinsic to the physical layer, such as wireless channel propagation, can be exploited to lock useful information. This information then can be automatically unlocked using real time analog RF means. In this paper retrodirective array, RDA, technology for spatial encryption in the multipath environment is for the first time combined with the directional modulation, DM, method normally associated with free space secure physical layer communications. We show that the RDA can be made to operate more securely by borrowing DM concepts and that the DM enhanced RDA arrangement is suitable for use in a multipath environment.
Resumo:
In this paper, we propose a new learning approach to Web data annotation, where a support vector machine-based multiclass classifier is trained to assign labels to data items. For data record extraction, a data section re-segmentation algorithm based on visual and content features is introduced to improve the performance of Web data record extraction. We have implemented the proposed approach and tested it with a large set of Web query result pages in different domains. Our experimental results show that our proposed approach is highly effective and efficient.
Resumo:
Balance maintenance relies on a complex interplay between many different sensory modalities. Although optimal multisensory processing is thought to decline with ageing, inefficient integration is particularly associated with falls in older adults. We investigated whether improved balance control, following a novel balance training intervention, was associated with more efficient multisensory integration in older adults, particularly those who have fallen in the past. Specifically, 76 healthy and fall-prone older adults were allocated to either a balance training programme conducted over 5 weeks or to a passive control condition. Balance training involved a VR display in which the on-screen position of a target object was controlled by shifts in postural balance on a Wii balance board. Susceptibility to the sound-induced flash illusion, before and after the intervention (or control condition), was used as a measure of multisensory function. Whilst balance and postural control improved for all participants assigned to the Intervention group, improved functional balance was correlated with more efficient multisensory processing in the fall-prone older adults only. Our findings add to growing evidence suggesting important links between balance control and multisensory interactions in the ageing brain and have implications for the development of interventions designed to reduce the risk of falls.
Resumo:
Bovine Respiratory Disease (BRD) is considered to be one of the most significant causes of economic loss in cattle worldwide. The disease has multifactorial aetiology, where viral induced respiratory damage can predispose animals to developing secondary bacterial infections. Accurate identification of viral infected animals prior to the onset of bacterial infection is necessary to reduce the overuse of antimicrobial treatments and minimize further economic losses from reduced production capacity and death. This research focuses on Bovine Parainfluenza Virus Type 3 (BPIV-3), one of the viruses involved in generating BRD. Vaccination measures for BPIV-3 can induce a level of immunity preventing disease progression, however, not all animals respond equally and immunization can complicate disease diagnosis. Alternative diagnostic approaches are required to identify animals which fail to respond to vaccination during infection outbreaks and are therefore likely to be more susceptible to secondary bacterial infections. Mass spectrometry based metabolomics was employed to identify plasma markers capable of differentiating between vaccinated and non-vaccinated calves after challenge with BPIV-3. Differentiation of vaccinated and non-vaccinated study groups (n=6) was possible as early as day 2 post-BPIV-3 challenge up until day 20 using a panel of potential metabolite markers. This study illustrates the potential for metabolomics to provide more detailed information on animal vaccination status that could be used to develop tools for improved herd health management, reduce economic loss through rapid identification and isolation of animals without immune protection (improving herd level immunity) and help reduce the usage of antimicrobial therapeutic treatments in animals.