Abstract The more embedded and ubiquitous computing devices are introduced in everyday domestic applications, the more important the aesthetics of interfaces become. In this paper, our interest is the aesthetics of domestic robots as an example of buddy-ing physical devices. We report on observations of using Sony AIBO dogs in outreach programs to new and untraditional user groups. The sessions were... The more embedded and ubiquitous computing devices are introduced in everyday domestic applications, the more important the aesthetics of interfaces become. In this paper, our interest is the aesthetics of domestic robots as an example of buddy-ing physical devices. We report on observations of using Sony AIBO dogs in outreach programs to new and untraditional user groups. The sessions were part of Narrative Lab workshops and Institute of Creative Technology (IOCT) Salon workshops. In the case of Narrative Lab workshop, AIBO robots were introduced to writers who were exploring new means of inspiration and writing in new media. A more elaborate study was conducted with undergraduate Humanities students. The observations in this paper will focus on the reactions towards the dog look and behaviour. It explores the relationships between the behaviours demonstrated, the physical appearance of the robot, and the users’ perceptions. We call upon aesthetics theories, aesthetics of interfaces, aesthetic metrics, and art theories research in our analysis of these observations. The outcome will be in the form of a framework to assess physical domestic devices design for buddy-ing interaction leading to design guidelines. Read More Read Less

Abstract Embedding Machine Learning technology into Medical Diagnosis Systems adds a new potential to such systems. This paper presents an algorithm using multilayer perceptron based Artificial Neural Network (ANN) for detecting swine flu accurately and efficiently. In this approach once an ANN model is trained using appropriate data patterns of different patients some having swine flu, the model becomes intelligent and... Embedding Machine Learning technology into Medical Diagnosis Systems adds a new potential to such systems. This paper presents an algorithm using multilayer perceptron based Artificial Neural Network (ANN) for detecting swine flu accurately and efficiently. In this approach once an ANN model is trained using appropriate data patterns of different patients some having swine flu, the model becomes intelligent and is ready for detecting correctly whether the case is a case of swine flu or not. Firstly feature selection is performed, which is driven by the goal of confirming a target class by a feature s discriminating power. A k-NN based algorithm is used to reduce the size of data to be used for training the ANN model with an objective of making the training more efficient and accurate. Then the structure of a pattern is represented by these important features along with a class label. Once the ANN model is trained using patterns of training set related to Swine flu, its performance is evaluated on test pattern set of Swine flu. The results achieved with the proposed approach demonstrate the ability of the algorithm to provide high level of accuracy for the Swine Flu classification problem. When the assessment (classification) ability of the ANN model is compared with that of CBR approaches, the superiority of the neural network approach is established. Read More Read Less

Abstract The benefit of random linear network coding (RLNC) has been studied in wired and wireless networks. Furthermore, it has been shown that RLNC improves throughput, bandwidth and robustness to network changes or link failures, howeverit is highly susceptible to errorpropagation. As existing error correcting codes are not adapted to such errors, recently Kötter and Kschischang introduced a new subspace code... The benefit of random linear network coding (RLNC) has been studied in wired and wireless networks. Furthermore, it has been shown that RLNC improves throughput, bandwidth and robustness to network changes or link failures, howeverit is highly susceptible to errorpropagation. As existing error correcting codes are not adapted to such errors, recently Kötter and Kschischang introduced a new subspace code in the context of a noncoherent transmission model. In this paper, we evaluate the performance of such code in the case of transmission over a network where each link is a q-ary Gilbert Elliot channel. We consider the conditional RLNC over Fq where, if all random coefficients are nulls, they have to be regenerated until at least one of them is not null. Simulation results verify theoretical results of [8] for different channel and coder parameters. Read More Read Less