Mixed feature selection based on granulation and approximation

Distance metric learning-based multi-granularity neighborhood rough sets for attribute reduction

Neighborhood rough set based heterogeneous feature subset selection

A novel approach to attribute reduction based on weighted neighborhood rough sets

Nowadays many applications generate mixed data objects consisting of numerical and categorical attributes. Simultaneously dealing with mixed objects is more challenging and various approaches convert one type to another one to face this issue. But in many cases this leads to information loss. Therefore integrating categorical and numerical attributes sounds reasonable since it keeps the original format of any attribute. In this paper we focus on clustering and especially density-based clustering as one of the well-known clustering approaches well-performed on arbitrary shape clusters. Density-based clustering algorithms require a distance measure to discover dense regions. Therefore we introduce the distance hierarchy as a distance measure appropriate for both categorical and numerical attributes. However setting the parameters regarding any parametric clustering algorithm could be another issue. Therefore we employ minimum description length principle to automate this process.

A neighborhood rough set model with nominal metric embedding

Neighborhood margin rough set: Self-tuning neighborhood threshold

Clustering algorithm for mixed datasets using density peaks and Self-Organizing Generative Adversarial Networks

This paper proposes to exploit heterogeneous data, i.e. data described by both numerical and categorical features, so as to gain knowledge about the categorical attributes from the numerical ones. More precisely, it aims at discovering whether, according to a given data set, based on information provided by the numerical attributes, some categorical attributes actually are ordinal ones and, additionally, at establishing ranking relations between the category values. To that aim, the paper proposes the 3-step methodology OSACA, standing for Order Seeking Algorithm for Categorical Attributes: it first consists in extracting gradual patterns from the numerical attributes, to identify rich ranking information about the data; it then applies mathematical morphology tools, more precisely alternated filters, to induce an associated order on the categorical attributes. The third step evaluates the quality of the candidate rankings through an original measure derived from the rank entropy discrimination.

The basic notions in rough set theory are lower and upper approximation operators defined by a fixed binary relation. This paper proposes an intuitionistic fuzzy rough set (IFRS) model which is a combination of intuitionistic fuzzy set (IFS) and rough set. We approximate an IFS by using soft binary relations instead of fixed binary relations. By using this technique, we get two pairs of intuitionistic fuzzy (IF) soft sets, called the upper approximation and lower approximation with respect to foresets and aftersets. Properties of newly defined rough set model (IFRS) are studied. Similarity relations between IFS with respect to this rough set model (IFRS) are also studied. Finally, an algorithm is constructed depending on these approximations of IFSs and score function for decision-making problems, although a method of decision-making algorithm has been introduced for fuzzy sets already. But, this new IFRS model is more accurate to solve the problem because IFS has degree of nonmembership and degree of hesitant.

Anonymized data publication has received considerable attention from the research community in recent years. For numerical sensitive attributes, most of the existing privacy preserving data publishing techniques concentrate on microdata with multiple categorical sensitive attributes or only one numerical sensitive attribute. However, many real-world applications may contain multiple numerical sensitive attributes. Directly applying the existing single-numerical-sensitive-attribute and multiple categorical-sensitive-attributes privacy preserving techniques often causes unexpected private information disclosure. They are particularly prone to the proximity breach, a privacy threat specific to numerical sensitive attributes in data publication. In this paper we propose a privacy-preserving data publishing method, namely MNSACM, that uses the ideas of clustering and Multi-Sensitive Bucketization (MSB) to publish microdata with multiple numerical sensitive attributes. Through an example we show the effectiveness of this method in privacy protection tomultiple numerical sensitive attributes.

A database can consist of numerical and non-numerical attributes. However, several data processing algorithms, such as K-means clustering, can be used only in a dataset with numerical attributes. Data generalization by using Naïve Bayes and K-means clustering methods is usually employed WEKA (Waikato environment for knowledge analysis) application. Although the strength of WEKA lies in increasingly complete and sophisticated algorithms, the success of data mining still lies in the knowledge factor of the human implementer. The task of collecting high-quality data and knowledge of modeling and the use of appropriate algorithms is needed to guarantee the accuracy of the expected formulations. In this paper, we propose a simple web-based application that can be used like WEKA. The methodology used in this study includes several stages. The first stage is the preparation of data, which is the tic-tac-toe game dataset that is converted to CSV (comma-separated values) format. The next stage is the process of modifying data from non-numeric to numeric, specifically for clustering with the K-means algorithm. Afterward, the calculation of the distance between data is conducted and followed by data clustering. The final stage is the summary of these processes and results. From the experimental results, it was found that clustering can be done on categorical attributes that are transformed first into the numerical form using web-based applications.

Mining quantitative association rules is an important topic of data mining since most real world databases have both numerical and categorical attributes. Typical solutions involve partitioning each numerical attribute into a set of disjoint intervals, interpreting each interval as an item, and applying standard boolean association rule mining. Commonly used partitioning methods construct set of intervals that either have equal width or equal cardinality. We introduce an adaptive partitioning method based on repeatedly merging smaller intervals into larger ones. This method provides an effective compromise between the equal width and equal cardinality criteria. Experimental results show that the proposed method is an effective method and improves on both equal-width partitioning and equal-cardinality partitioning.

Feature selection is one of the widely used pre-processing techniques to deal with large data sets. In this context, rough set theory has been successfully implemented for feature selection of discrete data set but in case of continuous data set it requires discretization, which may cause information loss. Fuzzy rough set theory approaches have also been used successfully to resolve this issue as it can handle continuous data directly. Moreover, almost all feature selection techniques are used to handle homogeneous data set. In this article, the center of attraction is on heterogeneous feature subset reduction. A novel intuitionistic fuzzy neighborhood models have been proposed by combining intuitionistic fuzzy sets and neighborhood rough set models by taking an appropriate pair of lower and upper approximations and generalize it for feature selection, supported with theory and its validation. An appropriate algorithm along with application to a data set has been added.

• Neural Network feature selection methodology for both numerical and categorical data. • Embedded feature selection provides model and feature subset in a single training. • Novel feature embedding for dealing with high-cardinality categorical features. • Validation of the methodology on several open and real industry dataset. • Comparison with several state-of-the-art feature selection methodologie. In an era of effortless data collection, the impact of machine learning — especially neural networks (NNs) — is undeniable. As datasets grow in size and complexity, efficiently handling mixed data types, including categorical and numerical features, becomes critical. Feature encoding and selection play a key role in improving NN performance, efficiency, interpretability, and generalisation. This paper presents GLEm-Net (Grouped Lasso with Embeddings Network), a novel NN-based approach that seamlessly integrates feature encoding and selection directly into the training process. GLEm-Net uses embedding layers to process categorical features with high cardinality, simplifying the model and improving generalisation. By extending the grouped Lasso regularisation to explicitly consider categorical features, GLEm-Net automatically identifies the most relevant features during training and returns them to the analyst. We evaluate GLEm-Net on open and proprietary industry datasets and compare it to state-of-the-art feature selection methodologies. Results show that GLEm-Net adapts to each dataset by allowing the NN to directly select subsets of most important features, offering on par performance with the best state-of-the-art feature selection methods, while eliminating the need for the external feature encoding and selection steps that are now incorporated in the NN training stage.

Neighborhood rough sets are an effective model for handling numerical and categorical data entangled with vagueness, imprecision, or uncertainty. However, existing neighborhood rough set models and their feature selection methods treat each sample equally, whereas different types of samples inherently play different roles in constructing neighborhood granules and evaluating the goodness of features. In this study, the sample weight information is first introduced into neighborhood rough sets, and a novel weighted neighborhood rough set model is consequently constructed. Then, considering the lack of sample weight information in practical data, a margin-based weight optimization function is designed, based on which a gradient descent algorithm is provided to adaptively learn sample weights through maximizing sample margins. Finally, an average granule margin measure is put forward for feature selection, and a forward-adding heuristic algorithm is developed to generate an optimal feature subset. The proposed method constructs the weighted neighborhood rough sets using sample weights for the first time and is able to yield compact feature subsets with a large margin. Extensive experiments and statistical analysis on UCI datasets show that the proposed method achieves highly competitive performance in terms of feature reduction rate and classification accuracy when compared with other state-of-the-art methods.