No abstract

No abstract

Numerous dimensionality-reducing representations of time series have been proposed in data mining and have proved to be useful, especially in handling a high volume of time series data. Among them, widely used symbolic representations such as symbolic aggregate approximation and piecewise aggregate approximation focus on information of local averages of time series. To compensate for such methods, several attempts were made to include trend information. However, the included trend information is quite simple, leading to great information loss. Such information is hardly extendable, so adjusting the level of simplicity to a higher complexity is difficult. In this paper, we propose a new symbolic representation method called transitional symbolic aggregate approximation that incorporates transitional information into symbolic aggregate approximations. We show that the proposed method, satisfying a lower bound of the Euclidean distance, is able to preserve meaningful information, including dynamic trend transitions in segmented time series, while still reducing dimensionality. We also show that this method is advantageous from theoretical aspects of interpretability, and practical and superior in terms of time-series classification tasks when compared with existing symbolic representation methods.

No abstract

Mobile learning has been a topic of research and development for 20 years. Over that time it has encompassed a wide range of concepts, theories, designs, experiments and evaluations. With increasing interest in the subject from researchers and practitioners, a comprehensive, yet accessible, overview of mobile learning that encompasses its many facets and features can provide a useful snapshot of the field. Although there have been many reviews of the mobile learning literature, these have tended to focus mostly on the nature of the work from a research perspective. Further, they have focused on a specific subset of the overall literature. Some review articles have, for example, confined themselves to a particular type of mobile learning, such as mobile language learning (e.g. Viberg & Gronlund, 2012). Others have focused on a specific conference series (e.g. Wingkvist & Ericsson, 2011) or subset of journals (e.g. Pollara &Broussard, 2011). In contrast, we have attempted to provide a more generic overview, and analyzed a large number of articles on mobile learning according to their main themes, concepts and concerns. These articles have come from a representative range of journals, books and conference proceedings. This poster provides a full-landscape view of the field of mobile learning in the form of a mind map. Whilst this approach to visualization is relatively subjective, it is a qualitative approach that allows us to find creative associations between ideas, as opposed to some other approaches that simply present quantitative data (Davies, 2011). It also potentially supports additional services such as certain types of information search (Beel &Gipp, 2010). Through an iterative process of refinement, we have applied the main analysis categories of research, learner, learning, content, design and technology, with a range of subcategories and representative exemplars. In addition, we contextualize the key developments in mobile learning in a timeline, in order to trace its evolution as a field of research.

No abstract

High level of accuracy in forecasting heat demand of each district is required for operating and managing the district heating efficiently. Heat demand has a close connection with the demands of the previous days and the temperature, general demand forecasting methods may be used forecast. However, there are some exceptional situations to apply general methods such as the exceptional low demand in weekends or vacation period. We introduce a new method to forecast the heat demand to overcome these situations, using the linearities between the demand and some other factors. Our method uses the temperature and the past 7 daysʼ demands as the factors which determine the future demand. The model consists of daily and hourly models which are multiple linear regression models. Appling these two models to historical data, we confirmed that our method can forecast the heat demand correctly with reasonable errors.

In the si mul ati on output anal ysi s, bootstrap method i s an appl i cabl e resampl i ng techni que to i nsuffi ci ent data whi ch are not si gni fi cant stati sti cal l y. The movi ng bl ock bootstrap, the stati onary bootstrap, and the threshol d bootstrap are typi cal bootstrap methods to be used for autocorrel ated ti me seri es data. They are nonparametri c methods for stati onary ti me seri es data, whi ch correctl y descri be the ori gi nal data. In the si mul ati on output anal ysi s, however, we may not use them because of the non-stati onari ty i n the data set caused by the trend such as i ncreasi ng or decreasi ng. In these cases, we can get ri d of the trend by di fferenci ng the data, whi ch guarantees the stati onari ty. We can get the bootstrapped data from the di fferenced stati onary data. Taki ng a reverse transform to the bootstrapped data, fi nal l y, we get the pseudo-sampl es for the ori gi nal data. In thi s paper, we i ntroduce the appl i cabi l i ty of bootstrap methods to the ti me seri es data havi ng trend, and then veri fy i t through the stati sti cal anal yses.