China
Africa
Explore chapters and articles related to this topic
Human-centred design
Published in Lisette van Gemert-Pijnen, Saskia M. Kelders, Hanneke Kip, Robbert Sanderman, eHealth Research, Theory and Development, 2018
The smartwatch that will be developed has to tightly reflect the needs of the user group consisting of seniors, since they may have very specific requirements. The interaction between user and technology should be as simple as possible because of the target group, so methods such as Cognitive Work or Task Analysis would be less suitable, because they include an analysis of already existing complex tasks which is not the case here. Therefore, these methods are more suitable for more complex technologies, such as a fitness tracker for athletes who are looking to develop complex training schedules or optimize diet or energy expenditures. The current type of design needed, with relatively simple interaction and a very specific context, would benefit most from methods as Personas, Contextual Inquiry and Participatory Design.
Context-Driven Image Analysis: Cognition Network Language
Published in Gerd Binnig, Ralf Huss, Günter Schmidt, Tissue Phenomics, 2018
This means learning or training does not necessarily require labeling or semantics. Today most NN procedures indeed use labeling but do not use semantics at all. The future of automatic data understanding lies probably in combining NNs with semantic processing and by making use of labeling as well as correlations and cluster analysis. This way a system could possibly be trained in a less labor-intensive fashion on complex problems than today. The different aspects have their individual pros and cons. Semantics is the way to train a system with what is known by humans. If something is already known, there is no need for complex training procedures. You tell the system similar to how you would tell another person. The combination of labeling or annotating with training of NNs represents a mix of using human knowledge combined with machine learning techniques. This is also a way how humans often communicate with each other, for example, by pointing at something and connecting a name to it. At Definiens, we have collected experiences over many years in transforming expert knowledge into the computer language CNL. This means translating human semantics into machine semantics. From this experience, we know that experts often do not know why they recognize something, but they definitely know that they indeed have recognized a relevant object. In this case, they explain what they see, categorize the findings, and simply point at the related regions or objects. In such a case where knowledge cannot be formulated semantically, annotating images and training the machine might be a good option.
Theories of aging and adaptation
Published in Peter G. Coleman, Ann O’Hanlon, Aging and Development, 2017
Peter G. Coleman, Ann O’Hanlon
Consequently, this model addresses criticisms put forward by Biggs (1999) among others, warning against the use of self-report assessments. It does, however, necessitate detailed and complex training. The complexity of the training is not necessarily a bad thing. Researchers always need to become increasingly detailed and refined in their observations of behaviour and experiences, and it should not be surprising that the achievement of such sophisticated insights does necessitate detailed training. We will describe further the value of this conceptual framework in describing research on attitudes to aging in the next chapter. Although offering very rich insights, analysis of the AAI is time consuming and complex. Areas for future research include developing and testing new self-report measures. This would save on time, and increase the use of this model in a wider array of settings.
Quiet Eye as a Mechanism for Table Tennis Performance under Fatigue and Complexity
Published in Journal of Motor Behavior, 2022
Andrada Vincze, Răzvan Jurchiș
Interested not only in the contribution of QE to performance but also in its functionality, we investigated whether it works as a mechanism to cope with more complex tasks (Klostermann et al., 2013). Accordingly, we expected that athletes will show longer QE in the more complex tasks. Contrary to previous results and to our expectations, we found no difference in QE, as a function of different levels of complexity. There is one plausible explanation for these results that comes from the structural aspects of the sample. As top performers, they are exposed to complex training tasks (identical to those that were tested in our study) on a daily basis. Thus, it is possible that the skills required to perform this task have been automatized and, therefore, they need less time to identify the relevant visual information in order to successfully hit the ball.
Variations in physiotherapy practices across reablement settings
Published in Physiotherapy Theory and Practice, 2020
Marianne Eliassen, Nils O. Henriksen, Siri Moe
In flexibly structured teams that reported sufficient PT resources, HTs stated in the interview that this type of intervention was exciting and educational. In contrast, an HT from a team with poor PT resources found it challenging to implement complex training interventions. One HT stated a preference for standardized exercises in some cases: “It’s not my subject in a way. I just have to do my best and listen to what she [the PT] has to say […] No, I don’t find it very easy. To evaluate what is the right exercise, no, that’s not my subject, really. That’s my experience. At least if there are any limitations, pain for example, if they are in pain or they feel that it takes… that they are exhausted afterwards, or… No. No. In some cases, we have these typical ‘Helbostad’ exercises, where we do those four exercises. Thirty on each, you know. That is more tangible. But in the more diffuse cases, it is not that simple, I think.” (HT)
Effect of AAC partner training using video on peers’ interpretation of the behaviors of presymbolic middle-schoolers with multiple disabilities*
Published in Augmentative and Alternative Communication, 2018
Christine Holyfield, Janice Light, Kathryn Drager, David McNaughton, Jessica Gormley
Despite recommendations from Kent-Walsh and McNaughton (2005), the current training did not include generalization of use of the skills targeted. Doing so might have bolstered the impact of the peer training. For instance, peers would have likely benefited from the opportunity to generalize their newly acquired skills by interpreting the behavior of students with multiple disabilities during face-to-face conversations. However, one important feature of the training used in the current study was its efficiency (under 15 min to complete). Adding more features to the training may increase its efficacy, but any increases in benefits from growing the training would have to be weighed against the additional time required to implement the more complex training. Alternatively, it may be possible to remove some aspects of the training in order to increase efficiency with minimal costs to effectiveness. For example, an online bank of video exemplars of communicative behaviors, along with the definitions and linguistic maps associated with those behaviors, could be made available for a wide range and a large number of communication partners to access and study independently, rather than requiring one-on-one time with a clinician. Future research could systematically manipulate specific components and content or compare different components to determine which approach to partner training is most effective in promoting accurate and consistent responsivity from communication partners.