scholarly journals CUR: Group Profiling with Community-based Users’ Representation

2018 ◽  
Author(s):  
João Emanoel Ambrósio Gomes ◽  
Ricardo B. C. Prudêncio ◽  
André C. A. Nascimento
Keyword(s):  
Feature Selection ◽  
Relevant Feature ◽  
Selection Methods ◽  
Community Based ◽  
Feature Sets ◽  
Network Community ◽  
Representation Method ◽  
Selection Algorithms ◽  
Content Information ◽  
Better Than

Group profiling methods aim to construct a descriptive profile for communities in social networks. Before the application of a profiling algorithm, it is necessary to collect and preprocess the users’ content information, i.e., to build a representation of each user in the network. Usually, existing group profiling strategies define the users’ representation by uniformly processing the entire content information in the network, and then, apply traditional feature selection methods over the user features in a group. However, such strategy may ignore specific characteristics of each group. This fact can lead to a limited representation for some communities, disregarding attributes which are relevant to the network perspective and describing more clearly a particular community despite the others. In this context, we propose the community-based user’s representation method (CUR). In this proposal, feature selection algorithms are applied over user features for each network community individually, aiming to assign relevant feature sets for each particular community. Such strategy will avoid the bias caused by larger communities on the overall user representation. Experiments were conducted in a co-authorship network to evaluate the CUR representation on different group profiling strategies and were assessed by hu- man evaluators. The results showed that profiles obtained after the application of the CUR module were better than the ones obtained by conventional users’ representation on an average of 76.54% of the evaluations.

scholarly journals Framework for the Ensemble of Feature Selection Methods

2021 ◽  
Vol 11 (17) ◽  
pp. 8122
Author(s):  
Maritza Mera-Gaona ◽  
Diego M. López ◽  
Rubiel Vargas-Canas ◽  
Ursula Neumann
Keyword(s):  
Machine Learning ◽  
Feature Selection ◽  
Selection Methods ◽  
Implementation Framework ◽  
Feature Sets ◽  
Machine Learning Classifiers ◽  
Selection Framework ◽  
The Stability ◽  
Selection Algorithms ◽  
Machine Learning Models

Feature selection (FS) has attracted the attention of many researchers in the last few years due to the increasing sizes of datasets, which contain hundreds or thousands of columns (features). Typically, not all columns represent relevant values. Consequently, the noise or irrelevant columns could confuse the algorithms, leading to a weak performance of machine learning models. Different FS algorithms have been proposed to analyze highly dimensional datasets and determine their subsets of relevant features to overcome this problem. However, very often, FS algorithms are biased by the data. Thus, methods for ensemble feature selection (EFS) algorithms have become an alternative to integrate the advantages of single FS algorithms and compensate for their disadvantages. The objective of this research is to propose a conceptual and implementation framework to understand the main concepts and relationships in the process of aggregating FS algorithms and to demonstrate how to address FS on datasets with high dimensionality. The proposed conceptual framework is validated by deriving an implementation framework, which incorporates a set of Phyton packages with functionalities to support the assembly of feature selection algorithms. The performance of the implementation framework was demonstrated in several experiments discovering relevant features in the Sonar, SPECTF, and WDBC datasets. The experiments contrasted the accuracy of two machine learning classifiers (decision tree and logistic regression), trained with subsets of features generated either by single FS algorithms or the set of features selected by the ensemble feature selection framework. We observed that for the three datasets used (Sonar, SPECTF, and WD), the highest precision percentages (86.95%, 74.73%, and 93.85%, respectively) were obtained when the classifiers were trained with the subset of features generated by our framework. Additionally, the stability of the feature sets generated using our ensemble method was evaluated. The results showed that the method achieved perfect stability for the three datasets used in the evaluation.

scholarly journals A Unified View of Causal and Non-causal Feature Selection

2021 ◽  
Vol 15 (4) ◽  
pp. 1-46
Author(s):  
Kui Yu ◽  
Lin Liu ◽  
Jiuyong Li
Keyword(s):  
Feature Selection ◽  
Bayesian Network ◽  
Synthetic Data ◽  
Selection Methods ◽  
Bayesian Network Model ◽  
Real World Data ◽  
Feature Sets ◽  
Unified View ◽  
Optimal Feature ◽  
Different Levels

In this article, we aim to develop a unified view of causal and non-causal feature selection methods. The unified view will fill in the gap in the research of the relation between the two types of methods. Based on the Bayesian network framework and information theory, we first show that causal and non-causal feature selection methods share the same objective. That is to find the Markov blanket of a class attribute, the theoretically optimal feature set for classification. We then examine the assumptions made by causal and non-causal feature selection methods when searching for the optimal feature set, and unify the assumptions by mapping them to the restrictions on the structure of the Bayesian network model of the studied problem. We further analyze in detail how the structural assumptions lead to the different levels of approximations employed by the methods in their search, which then result in the approximations in the feature sets found by the methods with respect to the optimal feature set. With the unified view, we can interpret the output of non-causal methods from a causal perspective and derive the error bounds of both types of methods. Finally, we present practical understanding of the relation between causal and non-causal methods using extensive experiments with synthetic data and various types of real-world data.

scholarly journals Boosting Atomic Orbit Search Using Dynamic-Based Learning for Feature Selection

Mathematics ◽  
2021 ◽  
Vol 9 (21) ◽  
pp. 2786
Author(s):  
Mohamed Abd Elaziz ◽  
Laith Abualigah ◽  
Dalia Yousri ◽  
Diego Oliva ◽  
Mohammed A. A. Al-Qaness ◽  
...  
Keyword(s):  
Feature Selection ◽  
Atomic Orbital ◽  
Critical Role ◽  
Machine Learning Algorithms ◽  
Selection Methods ◽  
Learning Technique ◽  
Real World Applications ◽  
Time And Space Complexity ◽  
Better Than ◽  
Search Domain

Feature selection (FS) is a well-known preprocess step in soft computing and machine learning algorithms. It plays a critical role in different real-world applications since it aims to determine the relevant features and remove other ones. This process (i.e., FS) reduces the time and space complexity of the learning technique used to handle the collected data. The feature selection methods based on metaheuristic (MH) techniques established their performance over all the conventional FS methods. So, in this paper, we presented a modified version of new MH techniques named Atomic Orbital Search (AOS) as FS technique. This is performed using the advances of dynamic opposite-based learning (DOL) strategy that is used to enhance the ability of AOS to explore the search domain. This is performed by increasing the diversity of the solutions during the searching process and updating the search domain. A set of eighteen datasets has been used to evaluate the efficiency of the developed FS approach, named AOSD, and the results of AOSD are compared with other MH methods. From the results, AOSD can reduce the number of features by preserving or increasing the classification accuracy better than other MH techniques.

scholarly journals Improved Feature-Selection Method Considering the Imbalance Problem in Text Categorization

2014 ◽  
Vol 2014 ◽  
pp. 1-17 ◽  
Author(s):  
Jieming Yang ◽  
Zhaoyang Qu ◽  
Zhiying Liu
Keyword(s):  
Feature Selection ◽  
Text Categorization ◽  
Information Gain ◽  
Feature Selection Method ◽  
Support Vector ◽  
Selection Methods ◽  
Document Collections ◽  
Imbalance Problem ◽  
Important Approach ◽  
Selection Algorithms

The filtering feature-selection algorithm is a kind of important approach to dimensionality reduction in the field of the text categorization. Most of filtering feature-selection algorithms evaluate the significance of a feature for category based on balanced dataset and do not consider the imbalance factor of dataset. In this paper, a new scheme was proposed, which can weaken the adverse effect caused by the imbalance factor in the corpus. We evaluated the improved versions of nine well-known feature-selection methods (Information Gain, Chi statistic, Document Frequency, Orthogonal Centroid Feature Selection, DIA association factor, Comprehensive Measurement Feature Selection, Deviation from Poisson Feature Selection, improved Gini index, and Mutual Information) using naïve Bayes and support vector machines on three benchmark document collections (20-Newsgroups, Reuters-21578, and WebKB). The experimental results show that the improved scheme can significantly enhance the performance of the feature-selection methods.

Enhanced Filter Feature Selection Methods for Arabic Text Categorization

2018 ◽  
Vol 8 (2) ◽  
pp. 1-24 ◽  
Author(s):  
Abdullah Saeed Ghareb ◽  
Azuraliza Abu Bakara ◽  
Qasem A. Al-Radaideh ◽  
Abdul Razak Hamdan
Keyword(s):  
Feature Selection ◽  
Text Categorization ◽  
Selection Process ◽  
High Dimensional Data ◽  
Relevant Information ◽  
High Dimensional ◽  
Arabic Text ◽  
Relevant Feature ◽  
Associative Classification ◽  
Selection Methods

The filtering of a large amount of data is an important process in data mining tasks, particularly for the categorization of unstructured high dimensional data. Therefore, a feature selection process is desired to reduce the space of high dimensional data into small relevant subset dimensions that represent the best features for text categorization. In this article, three enhanced filter feature selection methods, Category Relevant Feature Measure, Modified Category Discriminated Measure, and Odd Ratio2, are proposed. These methods combine the relevant information about features in both the inter- and intra-category. The effectiveness of the proposed methods with Naïve Bayes and associative classification is evaluated by traditional measures of text categorization, namely, macro-averaging of precision, recall, and F-measure. Experiments are conducted on three Arabic text datasets used for text categorization. The experimental results showed that the proposed methods are able to achieve better and comparable results when compared to 12 well known traditional methods.

scholarly journals Automatically evolving difficult benchmark feature selection datasets with genetic programming

2020 ◽  
Author(s):  
Andrew Lensen ◽  
Bing Xue ◽  
Mengjie Zhang
Keyword(s):  
Feature Selection ◽  
Genetic Programming ◽  
Superior Performance ◽  
Future Research ◽  
Selection Methods ◽  
Wide Range ◽  
Selection Algorithms ◽  
Different Characteristics

© 2018 Copyright held by the owner/author(s). There has been a wealth of feature selection algorithms proposed in recent years, each of which claims superior performance in turn. A wide range of datasets have been used to compare these algorithms, each with different characteristics and quantities of redundant and noisy features. Hence, it is very difficult to comprehensively and fairly compare these feature selection methods in order to find which are most robust and effective. In this work, we examine using Genetic Programming to automatically synthesise redundant features for augmenting existing datasets in order to more scientifically test feature selection performance. We develop a method for producing complex multi-variate redundancies, and present a novel and intuitive approach to ensuring a range of redundancy relationships are automatically created. The application of these augmented datasets to well-established feature selection algorithms shows a number of interesting and useful results and suggests promising directions for future research in this area.

scholarly journals The Impact of Feature Selection Methods for Classifying Arabic Textual Data

2019 ◽  
Vol 8 (4) ◽  
pp. 1333-1338
Keyword(s):  
Feature Selection ◽  
Text Classification ◽  
Information Gain ◽  
Feature Space ◽  
Support Vector ◽  
Selection Methods ◽  
K Nearest Neighbors ◽  
Chi Square ◽  
Selection Algorithms ◽  
The Impact

Text classification is a vital process due to the large volume of electronic articles. One of the drawbacks of text classification is the high dimensionality of feature space. Scholars developed several algorithms to choose relevant features from article text such as Chi-square (x2 ), Information Gain (IG), and Correlation (CFS). These algorithms have been investigated widely for English text, while studies for Arabic text are still limited. In this paper, we investigated four well-known algorithms: Support Vector Machines (SVMs), Naïve Bayes (NB), K-Nearest Neighbors (KNN), and Decision Tree against benchmark Arabic textual datasets, called Saudi Press Agency (SPA) to evaluate the impact of feature selection methods. Using the WEKA tool, we have experimented the application of the four mentioned classification algorithms with and without feature selection algorithms. The results provided clear evidence that the three feature selection methods often improves classification accuracy by eliminating irrelevant features.

scholarly journals Automatically evolving difficult benchmark feature selection datasets with genetic programming

2020 ◽  
Author(s):  
Andrew Lensen ◽  
Bing Xue ◽  
Mengjie Zhang
Keyword(s):  
Feature Selection ◽  
Genetic Programming ◽  
Superior Performance ◽  
Future Research ◽  
Selection Methods ◽  
Wide Range ◽  
Selection Algorithms ◽  
Different Characteristics

© 2018 Copyright held by the owner/author(s). There has been a wealth of feature selection algorithms proposed in recent years, each of which claims superior performance in turn. A wide range of datasets have been used to compare these algorithms, each with different characteristics and quantities of redundant and noisy features. Hence, it is very difficult to comprehensively and fairly compare these feature selection methods in order to find which are most robust and effective. In this work, we examine using Genetic Programming to automatically synthesise redundant features for augmenting existing datasets in order to more scientifically test feature selection performance. We develop a method for producing complex multi-variate redundancies, and present a novel and intuitive approach to ensuring a range of redundancy relationships are automatically created. The application of these augmented datasets to well-established feature selection algorithms shows a number of interesting and useful results and suggests promising directions for future research in this area.

scholarly journals THE IDENTIFICATION OF DETERMINANT PARAMETER IN FOREST FIRE BASED ON FEATURE SELECTION ALGORITHMS

SINERGI ◽  
2019 ◽  
Vol 23 (3) ◽  
pp. 184
Author(s):  
Devi Fitrianah ◽  
Hisyam Fahmi
Keyword(s):  
Feature Selection ◽  
Forest Fire ◽  
Feature Selection Method ◽  
Selection Method ◽  
Weather Data ◽  
Selection Methods ◽  
Cloud Coverage ◽  
Average Accuracy ◽  
Selection Algorithms

This research conducts studies of the use of the Sequential Forward Floating Selection (SFFS) Algorithm and Sequential Backward Floating Selection (SBFS) Algorithm as the feature selection algorithms in the Forest Fire case study. With the supporting data that become the features of the forest fire case, we obtained information regarding the kinds of features that are very significant and influential in the event of a forest fire. Data used are weather data and land coverage of each area where the forest fire occurs. Based on the existing data, ten features were included in selecting the features using both feature selection methods. The result of the Sequential Forward Floating Selection method shows that earth surface temperature is the most significant and influential feature in regards to forest fire, while, based on the result of the Sequential Backward Feature Selection method, cloud coverage, is the most significant. Referring to the results from a total of 100 tests, the average accuracy of the Sequential Forward Floating Selection method is 96.23%. It surpassed the 82.41% average accuracy percentage of the Sequential Backward Floating Selection method.

Sign in / Sign up

Export Citation Format

Share Document