Annamma Abraham

Comparison of Supervised and Unsupervised Learning Algorithms for Pattern Classification

This paper presents a comparative account of unsupervised and supervised learning models and their pattern classification evaluations as applied to the higher education scenario. Classification plays a vital role in machine based learning algorithms and in the present study, we found that, though the error back-propagation learning algorithm as provided by supervised learning model is very efficient for a number of non-linear real-time problems, KSOM of unsupervised learning model, offers efficient solution and classification in the present study.

Rayleigh–Benard convection in a micropolar ferromagnetic fluid

Abstract The problem of Rayleigh–Benard convection in a micropolar ferromagnetic fluid layer permeated by a uniform, vertical magnetic field is investigated analytically with free–free, isothermal, spin-vanishing, magnetic boundaries. The influence of the various micropolar and magnetization parameters on the onset of stationary convection has been analysed. It is observed that the micropolar ferromagnetic fluid layer heated from below is more stable as compared with the classical Newtonian ferromagnetic fluid. The nature of influence of the magnetization parameters on convection in the micropolar ferromagnetic fluid is similar to that in the case of Newtonian ferromagnetic fluids. The influence of the micropolar parameters on convection in the ferromagnetic case is akin to its role in the non-magnetic fluid case. The critical wave number is found to be insensitive to the changes in the micropolar fluid parameters, but sensitive to the magnetization parameters.

Unsupervised Control Paradigm for Performance Evaluation

Intelligent control denotes the capacity to acquire and apply knowledge in control process. The important characteristics of intelligent control systems are information abstraction and knowledge-based decision making. There are different control paradigms available in the literature including Artificial Neural Networks, Fuzzy Logic Systems, Genetic Algorithms, Hybrid Models and others. This paper attempts to design open loop controller using Self Organizing Map and studies its nature and accuracy with an example.

Cluster Analysis: An Experimental Study

Clustering plays a vital role in machine based learning algorithms and in the present study, it is found that, the competitive learning algorithm that is very efficient for a number of non-linear real-time problems, offers efficient solution for clustering. This paper presents a comparative account of self-organizing models and proposes a hybrid selforganizing model for cluster analysis. The potential usefulness of cluster analysis for higher education scenario is taken to study in this paper.

An Adapter based Network Management Layer Framework for provisioning Telecom Services

The provisioning of telecom services requires the subscribers and service related information to be provisioned onto the network elements. The network elements or element management systems store the information received and is used for activation of the services for the subscribers. These telecom services could be wire-line or wireless based services. All the telecom service providers are rapidly shifting to Next Generation Networks (NGN) which are IP based packet networks. NGNs are independent of the underlying transport technology, hence it does not matter whether the underlying hardware technology is wire-line or wireless based. This enables the service providers to focus on defining new telecom services which could be delivered to customers. Thus, there arises competition between service providers in creation and provision of services to customers as quickly as possible. Also, quick customization of services that are already provided to the customers based on their needs is a major requirement. Hence, the challenge of delivering services by provisioning them quickly on the network is to be addressed as the users scale in numbers. As a part of addressing this challenge, there is a need for effective communication with the network elements for provisioning. A framework that can integrate with multiple Service Order systems, transform their requests into network commands and techniques used to communicate efficiently is mentioned. The commands are generated suitably for the network element and communicated to it based on SOAP protocol. It can work with network elements that expose an interface using Web Service Description Language (WSDL).