![]() ![]() The progenitor of our modern-day “weird” originally meant something closer to fate or destiny, with a little twist of the supernatural-a perfect description of the forces that bond them. ![]() That’s why the Old English word wyrd might be the best term to describe the origins of his beautiful friendship and subsequent collaboration with actor and musician Johnny Flynn. Robert Macfarlane, high chieftan of logofiles everywhere, has dedicated entire books, including 2015’s Landmarks and 2017’s The Lost Words, which he co-authored with Jackie Morris, to preserving words and their original meanings. It starts by eliciting the DW requirements and then. In fact, the proposal is an MDA-oriented UML profiles. In this paper, we describe how to apply MDA throughout the entire development process of DW. The MDA aims to automate the software engineering process, reducing thus the software development cost and improving its productivity. On the other hand, the Model Driven Architecture (MDA) is a standard approach which intended to support the whole phases of software manufacturing, by promoting the usage of models and transformations between them up to code generation. ![]() Generally, most of those approaches are interested in a particular aspect of DW (Storage, ETL processing, OLAP analysis, reporting, etc.) and don't cover its whole life-cycle. However, there are neither standard method that addresses the design of all DW layers nor a software process prescribed for this kind of engineering domain. Nowadays, many approaches have been proposed to design and develop Data Warehouse (DW). A comparison to three existing approaches shows that the proposed meta-classifier is competitive according to hamming-loss evaluation measure, and it is the most stable classifier according to hamming-loss standard deviation. The studied compromise is analysed according to its impact on the classifier complexity and on hamming-loss evaluation measure. Firstly, no predefined structure is imposed for learning label relations, and secondly, the meta-classifier is based on three measures giving control on the studied compromise. In this paper, a new meta-classifier with two main advantages is proposed for GMLC. This paper is motivated by the lack of a study analysing the compromise between handling label relations and limiting error propagation in GMLC, and by the fact that there is no known approach giving a control on that compromise to allow such a study. Most of existing approaches either ignore label relations, or can learn only relations fitting a predefined imposed structure. Ignoring those relations can lead to inconsistent predictions, but if they are considered, then a prediction error for one label will be propagated to all related labels. For example, in a movie catalog web page, a five stars action movie should be at least a one star suspense movie. In graded multi-label classification (GMLC), each data can be assigned to multiple labels according to a degree of membership on an ordinal scale, and with respect to label relations. ![]()
0 Comments
Leave a Reply. |