eulogize 发表于 2025-4-1 03:49:11
The Impact of Local Data Characteristics on Learning from Imbalanced Dataributions of the minority class and show that they could be approximated by analysing the neighbourhood of the learning examples from the minority class. We claim that the results of this analysis could be a basis for developing new algorithms. In this paper we show such possibilities by discussingGUISE 发表于 2025-4-1 08:20:00
The Relational Construction of Conceptual Patterns - Tools, Implementation and Theory compare different mathematical notions and tools used in qualitative data analysis. Moreover, since relations are representable by Boolean matrices, computing the conceptual-oriented operators is straightforward. Finally, the relational-based approach makes it possible to conceptually analyse not o金桌活画面 发表于 2025-4-1 12:53:18
SQL-Based KDD with Infobright’s RDBMS: Attributes, Reducts, Treesd inducing decision trees. We focus particularly on attribute reduction, which is important especially for high-dimensional data sets. The main technical contribution of this paper is a complete framework for calculating short reducts using SQL queries on data stored in a relational form, without a