A comparative dimensionality reduction study in telecom customer segmentation using deep learning and PCA

Abstract

Due to the increased competition between telecommunication operators and growing customers' churn rate, telecommunication companies were seeking to improve customer loyalty. In order to increase customer satisfaction, most telecom companies resort to customer segmentation which entails separating the targeted customers into different groups based on demographics or usage perspective including gender, age-group, buying behavior, usage pattern, special interests and other features that represent the customer. Customer segmentation adopted by most telecom operators to provide the customer with the right offer.

An Accurate, Robust and Low Dimensionality Deep Learning Localization Approach in DM-MIMO Systems Based on RSS

“…In [17] and [18], the Gaussian process regression (GPR) ML algorithm is employed based on RSS measurements in DM-MIMO systems. In [3], the performance of several ML algorithms, which are used in conjunction with fingerprint-based MT localization for DM-MIMO wireless systems configurations, is investigated and evaluated. In [9], RSS-based positioning using a machine learning method relies on the affinity propagation clustering algorithm and the GPR algorithm.…”

“…Location information has a great application potential in industry, medicine, emergency management, surveillance, controlling autonomous vehicles, and many other various fields [2]. At the same time, the demand for recognition of a mobile terminal's (MT) location has greatly increased so that numerous researches have been conducted on MTs' location [3]. Therefore, developping localization technology is becoming more and more important.…”

“…Therefore, developping localization technology is becoming more and more important. The source of most localization systems are based in urban settings [3]. Today, the global positioning system (GPS) is the most used technology for outdoor localization due to its availability [4].…”

Deep learning from a statistical perspective

“…However, in the standard autoencoders, there is no such ordering and orthogonality, which makes it more difficult to guarantee that the bases of the latent spaces are independent and the size of the latent spaces needs to be predetermined. For more detailed discussions regarding the relationship between t autoencoders and PCA, we refer the readers to Ladjal et al (2019) and Alkhayrat, Aljnidi, and Aljoumaa (2020).…”

Single- and Multi-Distribution Dimensionality Reduction Approaches for a Better Data Structure Capturing

“…Dimensionality reduction is a process of converting data from a high dimensional space to a lower dimensional space with the aim of preserving meaningful information from the original data. Dimensionality reduction can be applied in any field that has high dimensional data (a large number of variables) such as signal processing [1], speech recognition [2,3], neuroinformatics [4,5], bioinformatics [6,7], social media [8,9], telecoms [10], and computer vision [11], for data visualization, data exploration, noise reduction or as a pre-processing step to support classification models. An appropriate dimensionality reduction technique is related to the goodness of preserving the geometry (structure) of the data of interest.…”