Autonomous Market Segments Estimation Using Density Conscious Artificial Immune System Learner

Autonomous Market Segments Estimation Using Density Conscious Artificial Immune System Learner

Vishwambhar Pathak (BIT MESRA Jaipur, India)
DOI: 10.4018/978-1-5225-2234-8.ch006
OnDemand PDF Download:
List Price: $37.50


Automated exploration of groups of customers to understand customer behavior from raw data is highly required to support strategic decision making given the pressure of competitive market. Several mathematical and statistical methods have been applied for autonomous model estimation from multivariate data. The current paper investigates employability of new generation of bio-inspired metaheuristic algorithms, named the artificial immune system (AIS), which in the current proposition, learn through density based kernels. As such the model simulates probabilistic behavior of the dendritic cells (DCs) during recognition of the antigens and danger signals, whose learning has been modeled with an infinite Gaussian mixture model. The unsupervised learning capability of the model has been found to be effective for multivariate data.
Chapter Preview


Market segmentation refers to the marketing strategy that seeks to divide a broad market perspective into subsets of consumers, locations, and businesses, in terms of common needs, interests, and priorities in a way to define their strategies to target potential customers. It is more so sought by small scale companies to optimize the cost to increase consumer penetration. The analysis covers a wide range including geographic segmentation, demographic segmentation, behavioral segmentation, psychographic segmentation, occasional segmentation, segmentation by benefits, emotive Segmentation, cultural segmentation, and multi-variable account segmentation. Computational tools have been proved of great help in the recent knowledge world, where accurate data analyses have been made possible due to huge amount of data about business operations being stored continuously. Several mathematical and statistical methods have been developed and successfully applied help the decision makers. Data mining and machine learning algorithms of wide range including the Decision Trees, Classification and Regression Trees, Rough Sets, Self organizing maps, Fuzzy inference engines, and vast range of bio-inspired algorithms have been reported in research literature (Prabha & Ilango, 2014; Mattila, 2008; Yao, 2013; Taylor, n.d.; Chulis, 2012); majority of which have been implemented in successful commercial tools for the task. The current work focuses on a bio-inspired algorithm based on natural immune system dynamics modeled using density based model estimation methods, applied to extracting segments from multivariate data.

Complete Chapter List

Search this Book: