Search

In this paper we present one way of modeling candlestick patterns using interpolative Boolean algebra (IBA). This method shows a degree of fulfillment for observed patterns, thus giving traders easy interpretation of by how much candlesticks fit into different patterns. Candlestick patterns have been used for financial forecasting for a couple of decades on Western markets and they have become a mainstream trader's tool. Since the need for automated candlestick patterns discovery arose, some papers proposed fuzzy approach as a solution. Our decision to use IBA for modeling candlestick patterns comes from the fact that fuzzy logic has it limits and cannot be applied to these models. Proposed method is another approach to the same problem, but it could not be modeled using conventional fuzzy logic, because it is necessary for it to be in the Boolean frame. Results obtained from our tests are satisfactory and also open the opportunity for combining this technique with existing ones.

In this paper, we present a problem of candidates ranking for Master of Science studies at Faculty of Organizational Sciences, Belgrade (FOS). Current ranking model is based on weighted sum of two factors: average grade gained at undergraduate studies and number of points scored at the entrance exam for master studies. This universal model, although widespread and frequently used, is not entirely appropriate for a number of different Master of Science programs at FOS. Major problems of current model are that the model does not emphasize essential knowledge in accordance to wanted program and that the weighted sum is unable to model connection between variables. This paper presents a series of models, each more complex than the previous one, which are aggregating more relevant factor for ranking. Logical Aggregation (LA) is used as a method for the aggregation of certain variables in the last few models. LA is based on Interpolative Boolean algebra (IBA), a consistent multi-valued realization of Boolean algebra. This paper may be particularly interesting to all of those dealing with any form of students ranking, especially to university departments involved in enrolling students and selection for scholarships.