Pub. online:29 Jan 2021Type:Research ArticleOpen Access
Journal:Informatica
Volume 32, Issue 1 (2021), pp. 85–118
Abstract
The data-driven approach is popular to automate learning of fuzzy rules and tuning membership function parameters in fuzzy inference systems (FIS) development. However, researchers highlight different challenges and issues of this FIS development because of its complexity. This paper evaluates the current state of the art of FIS development complexity issues in Computer Science, Software Engineering and Information Systems, specifically: 1) What complexity issues exist in the context of developing FIS? 2) Is it possible to systematize existing solutions of identified complexity issues? We have conducted a hybrid systematic literature review combined with a systematic mapping study that includes keyword map to address these questions. This review has identified the main FIS development complexity issues that practitioners should consider when developing FIS. The paper also proposes a framework of complexity issues and their possible solutions in FIS development.
Journal:Informatica
Volume 21, Issue 1 (2010), pp. 13–30
Abstract
The genetic information in cells is stored in DNA sequences, represented by a string of four letters, each corresponding to a definite type of nucleotides. Genomic DNA sequences are very abundant in periodic patterns, which play important biological roles. The complexity of genetic sequences can be estimated using the information-theoretic methods. Low complexity regions are of particular interest to genome researchers, because they indicate to sequence repeats and patterns. In this paper, the complexity of genetic sequences is estimated using Shannon entropy, Rényi entropy and relative Kolmogorov complexity. The structural complexity based on periodicities is analyzed using the autocorrelation function and time delayed mutual information. As a case study, we analyze human 22nd chromosome and identify 3 and 49 bp periodicities.
Journal:Informatica
Volume 7, Issue 1 (1996), pp. 15–26
Abstract
In this paper, we propose to present the direct form recursive digital filter as a state space filter. Then, we apply a look-ahead technique and derive a pipelined equation for block output computation. In addition, we study the stability and multiplication complexity of the proposed pipelined-block implementation and compare with complexities of other methods. An algorithm is derived for the iterative computation of pipelined-block matrices.
Journal:Informatica
Volume 6, Issue 2 (1995), pp. 193–224
Abstract
We apply some concepts of Information-Based Complexity (IBC) to global and discrete optimization. We assume that only partial information on the objective is available. We gather this partial information by observations. We use the traditional IBC definitions and notions while defining formal aspects of the problem. We use the Bayesian framework to consider less formal aspects, such as expert knowledge and heuristics.
We extend the traditional Bayesian Approach (BA) including heuristics. We call that a Bayesian Heuristic Approach (BHA).
We discuss how to overcome the computational difficulties using parallel computing. We illustrate the theoretical concepts by three examples: by discrete problems of flow-shop scheduling and parameter grouping, and by a continuous problem of batch operations scheduling.
Journal:Informatica
Volume 3, Issue 3 (1992), pp. 301–337
Abstract
Small training sample effects common in statistical classification and artificial neural network classifier design are discussed. A review of known small sample results are presented, and peaking phenomena related to the increase in the number of features and the number of neurons is discussed.