Τμήμα Διοίκησης Επιχειρήσεων (Τεχνικές Αναφορές)
Permanent URI for this collection
Browse
Recent Submissions
Item Open Access Τα κρούσματα κοροναϊού COVID-19(Αυτο-έκδοση) Ανδρουλάκης, Γεώργιος; Λισγάρα, Ελένη; Ανδρουλάκης, Σίλας; Καρολίδης, Γεώργιος; Androulakis, George; Lisgara, Eleni; Androulakis, Silas; Karolidis, GeorgeThe number of the daily COVID-19 cases* comprises a time series, and time series forecasting is the act of modeling to predict its future values based on previous data. Although most forecasting methodologies focus on value-oriented results, local optima, there is some research on time-oriented methods that focus on predicting the time that the local optima will occur (Androulakis & Lisgara, 2007; Lisgara, Karolidis, & Androulakis, 2010a, 2010b, 2012). These techniques regard time series as an objective function subject to the factors affecting its values and apply nonlinear optimization techniques to approach the time that future minimum/maximum will happen. In this work, the technique provided by Lisgara, Karolidis & Androulakis, (Lisgara et al., 2010b) is exploited to predict the timing the COVID-19 time series will reach its maximum. Then, value-oriented methods are applied to extract an approximation of the expected number of cases until the predicted time period. For this study were used all avalable data until March the 21st (Dong, Du, & Gardner, 2020). The COVID-19 time series consists of the daily number of cases; at the beginning, its rate seems to grow exponentially, then the rate gradually slows down and merely the time series reaches its maximum value. Similarly, the rate decreases exponentially and eventually the series reaches its end-time (zero cases). The rate’s path is defined as the cycle of the phenomenon. It was observed that the cycle’s time extension would result to a lower maximum value, meaning less daily cases and, consequently, less daily serious cases. Also, it was also observed that such an extension would result to delaying the outbreak’s peak, meaning that the cases number would spread in a longer period of time. However, timing-control is vital to ensure that the healthcare system would not exceed its capacity.Item Open Access University of Patras students - Questionnaire Dropout May 2019(Department of Business Administration) -; -UPatrasStudentsMay2019Item Open Access Improvement of similarity-diversity trade-off in recommender systems based on a facility location modelPanteli, Antiopi; Boutsinas, Basilis; Παντελή, Αντιόπη; Βουτσινάς, ΒασίλειοςThere is a growing interest in the offering of novel alternative choices to users of recommender systems. These recommendations should match the target query while at the same time they should be diverse with each other in order to provide useful alternatives to the user, i.e. novel recommendations. In this paper, the problem of extracting novel recommendations, under the similarity-diversity trade-off, is modeled as a facility location problem. We formulate this trade-off as a multiple p-median problem solved by using biclustering. The results from tests in the benchmark Travel Case Base were satisfactory when compared to well-known recommender techniques, in terms of both similarity and diversity. Moreover, the experimental tests have shown that the proposed method is flexible enough, since a parameter of the adopted facility location model constitutes a regulator for the trade-off between similarity and diversity.Item Open Access Biclustering based on association analysis(2013-03-27) Βουτσινάς, Βασίλης; Boutsinas, BasilisClustering has been applied in a wide variety of disciplines and has also been utilized in many scientific areas. Usually, clustering algorithms construct either clusters of rows or clusters of columns of the input data matrix. Biclustering is a methodology where biclusters are formed by both a subset of rows and a subset of columns, such that objects represented by the first are the most similar to each other when compared over the latter. In this paper, we introduce a new biclustering technique, based on association rule mining, which can support different well-known biclustering models proposed in the literature. Experimental tests demonstrate the accuracy and efficiency of the proposed technique with respect to well known related ones.
