Abstract—This paper is a precipitate of handpicked research papers in the field of Information fusion specific to multiple classifier systems, following systematic consolidation, clearly spotlighting technology trends, issues, opportunities and challenges, apart from inking authors’ views and critical suggestions in the form of questions at the end of this paper. Large amounts of data are nowadays available because gathering data is easy and inexpensive. Most data is raw and to be useful relevant knowledge has to be extracted from it. Technology alone does not deliver a solution. Data mining is a process that must be reliable, repeatable with even with a little knowledge about data mining by the people it provides the solution. Data analytics is the day in and day out activity for almost all the organizations. People with business problem and with the data on that issue, expect more sophisticated and actionable information with higher accuracy and time (cost) effective solution to make decisions from the data. Data mining functionalities provide the organization with some form of intelligence. If we have several sets of data obtained from various sources, where the nature of features are different (heterogeneous features), a single classifier cannot be used to learn the information contained in all of the data. And if we have to perform different tests then each test generates data with a different number and type of features, which cannot be used collectively to train a single classifier. In such cases, data from each testing modality can be used to train a different classifier, whose outputs can then be combined. Applications in which data from different sources are combined to make a more informed decision are referred to as Information fusion applications, and ensemble based approaches like bagging and all the variants of boosting etc. have successfully been used for such applications. We compared the performance and performed the survey of these methods on a collection of machine-learning benchmarks. This paper reports the general survey and the results of applying Information Fusion (Multiple Classifier System) methods to a system that learns from various models, problems that may arise in implementing these models (algorithms) are explored and the research issues where the further work is expected to be done.
Index Terms—Information Fusion, Multiple Classifier, Machine Learning, Ensemble, Diversity, Data Mining, bagging, boosting.
1 * Associate Professor, CE-IT, Faculty of Engineering & Technology
** Dean, Faculty of Engineering & Technology, Charotar University of Science & Technology, Changa-388421, Anand, Gujarat (INDIA)
Cite: Amit P Ganatra and Yogesh P Kosta, "Trend Spotting Technological advancement, Issues, Challenges and Opportunities in Information Fusion Methods specific to Multiple Classifier Systems - A Survey Paper," International Journal of Computer Theory and Engineering vol. 2, no. 6, pp. 925-930, 2010.