We will review some of the major results in random graphs and some of the more challenging open problems. We will cover algorithmic and structural questions. We will touch on newer models, including those related to the WWW.

Random graphs

An overview of the self-organizing map algorithm, on which the papers in this issue are based, is presented in this article.

The Self-Organizing Map

Feature selection, as a data preprocessing strategy, has been proven to be effective and efficient in preparing data (especially high-dimensional data) for various data-mining and machine-learning problems. The objectives of feature selection include building simpler and more comprehensible models, improving data-mining performance, and preparing clean, understandable data. The recent proliferation of big data has presented some substantial challenges and opportunities to feature selection. In this survey, we provide a comprehensive and structured overview of recent advances in feature selection research. Motivated by current challenges and opportunities in the era of big data, we revisit feature selection research from a data perspective and review representative feature selection algorithms for conventional data, structured data, heterogeneous data and streaming data. Methodologically, to emphasize the differences and similarities of most existing feature selection algorithms for conventional data, we categorize them into four main groups: similarity-based, information-theoretical-based, sparse-learning-based, and statistical-based methods. To facilitate and promote the research in this community, we also present an open source feature selection repository that consists of most of the popular feature selection algorithms (http://featureselection.asu.edu/). Also, we use it as an example to show how to evaluate feature selection algorithms. At the end of the survey, we present a discussion about some open problems and challenges that require more attention in future research.

https://dl.acm.org/doi/pdf/10.1145/3136625

Feature Selection: A Data Perspective

Many ``real-world'' networks are clearly defined while most ``social'' networks are to some extent subjective. Indeed, the accuracy of empirically-determined social networks is a question of some concern because individuals may have distinct perceptions of what constitutes a social link. One unambiguous type of connection is sexual contact. Here we analyze data on the sexual behavior of a random sample of individuals, and find that the cumulative distributions of the number of sexual partners during the twelve months prior to the survey decays as a power law with similar exponents $\alpha \approx 2.4$ for females and males. The scale-free nature of the web of human sexual contacts suggests that strategic interventions aimed at preventing the spread of sexually-transmitted diseases may be the most efficient approach.

The Web of Human Sexual Contacts

Feature selection in machine learning: A new perspective

Recently, feature selection and dimensionality reduction have become fundamental tools for many data mining tasks, especially for processing high-dimensional data such as gene expression microarray data. Gene expression microarray data comprises up to hundreds of thousands of features with relatively small sample size. Because learning algorithms usually do not work well with this kind of data, a challenge to reduce the data dimensionality arises. A huge number of gene selection are applied to select a subset of relevant features for model construction and to seek for better cancer classification performance. This paper presents the basic taxonomy of feature selection, and also reviews the state-of-the-art gene selection methods by grouping the literatures into three categories: supervised, unsupervised, and semi-supervised. The comparison of experimental results on top 5 representative gene expression datasets indicates that the classification accuracy of unsupervised and semi-supervised feature selection is competitive with supervised feature selection.

Supervised, Unsupervised, and Semi-Supervised Feature Selection: A Review on Gene Selection

A convergent algorithm for orthogonal nonnegative matrix factorization

The Tikhonov regularized nonnegative matrix factorization (TNMF) is an NMF objective function that enforces smoothness on the computed solutions, and has been successfully applied to many problem domains including text mining, spectral data analysis, and cancer clustering. There is, however, an issue that is still insufficiently addressed in the development of TNMF algorithms, i.e., how to develop mechanisms that can learn the regularization parameters directly from the data sets. The common approach is to use fixed values based on a priori knowledge about the problem domains. However, from the linear inverse problems study it is known that the quality of the solutions of the Tikhonov regularized least square problems depends heavily on the choosing of appropriate regularization parameters. Since least squares are the building blocks of the NMF, it can be expected that similar situation also applies to the NMF. In this paper, we propose two formulas to automatically learn the regularization parameters from the data set based on the L-curve approach. We also develop a convergent algorithm for the TNMF based on the additive update rules. Finally, we demonstrate the use of the proposed algorithm in cancer clustering tasks.

Nonparametric tikhonov regularized NMF and its application in cancer clustering

This article describes how the use of Web 2.0 technologies in the field of learning is on the rise. By their nature, Web 2.0 technologies increase the interactivity between users where interactivity is considered to be a key to success in traditional classrooms. This article reviews recent studies in the field of Web 2.0 technologies for learning and their impacts on the learning experiences and investigates relationship between Web 2.0 technologies and pedagogy in higher education on student learning. Key findings about the impacts of using social networks like Facebook, Twitter, blogs and wikis on learning experiences are also discussed. Web 2.0 technologies' characteristics and the rationale of Web 2.0 technologies in learning will also be explored.

The Impact of Web 2.0 Technologies on the Learning Experience of Students in Higher Education: A Review.

Blind source separation (BSS) is a problem of recovering source signals from signal mixtures without or very limited information about the sources and the mixing process. From literatures, nonnegative matrix factorization (NMF) and independent component analysis (ICA) seem to be the mainstream techniques for solving the BSS problems. Even though the using of NMF and ICA for BSS is well studied, there is still a lack of works that compare the performances of these techniques. Moreover, the nonuniqueness property of NMF is rarely mentioned even though this property actually can make the reconstructed signals vary significantly, and thus introduces the difficulty on how to choose the representative reconstructions from several possible outcomes. In this paper, we compare the performances of NMF and ICA as BSS methods using some standard NMF and ICA algorithms, and point out the difficulty in choosing the representative reconstructions originated from the nonuniqueness property of NMF.

Andri Mirzal

Papers

Supervised, Unsupervised, and Semi-Supervised Feature Selection: A Review on Gene Selection

A convergent algorithm for orthogonal nonnegative matrix factorization

Nonparametric tikhonov regularized NMF and its application in cancer clustering

The Impact of Web 2.0 Technologies on the Learning Experience of Students in Higher Education: A Review.

NMF versus ICA for blind source separation