Kuruvilla Mathew

A GENETIC ALGORITHM ANALYSIS TOWARDS OPTIMIZATION SOLUTIONS

In today’s world, an optimal and intelligent problem solving approaches are required in every field, regardless of simple or complex problems. Researches and developers are trying to make machines and softwares more efficient and intelligent. This is where the Artificial Intelligence plays its role in developing efficient and optimal searching algorithm solutions. Genetic algorithm is one of most pervasive and advanced developed heuristic search technique in AI. Genetic algorithm (GA) is developed to find the most optimized solution for a given problem based on inheritance, mutation, selection and some other techniques. It was proved that genetic algorithms are the most powerful unbiased optimization techniques for sampling a large solution space. In this paper, we have used GA for the image optimization and Knapsack Problems, which are commonly found in a real world scenario. Furthermore, a research based on a tool that uses Genetic Algorithm, called the GA Playground is done to demonstrate the capability of solving the Knapsack Problem with the fitness function and a case study on how images can be reproduced using the optimal parameters. Lastly, a few methods such as the Hash Table and the Taguchi Method are suggested to improve the performance of the Genetic Algorithm.

Intelligent spam classification for mobile text message

This paper analyses the methods of intelligent spam filtering techniques in the SMS (Short Message Service) text paradigm, in the context of mobile text message spam. The unique characteristics of the SMS contents are indicative of the fact that all approaches may not be equally effective or efficient. This paper compares some of the popular spam filtering techniques on a publically available SMS spam corpus, to identify the methods that work best in the SMS text context. This can give hints on optimized spam detection for mobile text messages.

Ubiquitous Text Transfer Using Sound a Zero-Infrastructure Alternative for Simple Text Communication

Even in these days where data networks has increased much in terms of speed, bandwidth and penetration, the need for a low power, low bandwidth, ubiquitous networks is more pronounced than ever before. As the devices get smaller, their power supply is also limited, in according to the definition of “dust”, “skin” and “clay” in the ubiquitous computing paradigm. The possibility of these devices to be present in real world depends a lot on the key capability they must possess, which is to be network enabled, ubiquitously. This paper looks at the possibility of using the ever present signal “sound” as a ubiquitous medium of communication. We are currently experimenting on various possibilities and protocols that can make use of sound for text transmission between two electronic devices and this paper looks at some attempts in this direction. The initial phase of the experiment was conducted using a very large spectrum and encoding the entire ASCII text over audible sound spectrum. This gave a very large spectrum spread requirement which a very narrow frequency gap. The experimental results showed good improvement when the frequency gap was increased.

Ambient noise analysis on sound for use in wireless digital transmission

Ambient noise is present everywhere in varying degrees of frequencies and amplitudes. This paper tries to analyse the ambient noises in various environments ranging from quiet room to busy public places to remote jungle type locations in order to study the impact of noise while using sound as a low frequency wireless carrier for digital data. We try to find the suitable frequency band for text transmission using sound by identifying the sound spectrum band least susceptible to ambient noise and do some test communication.

Evaluation of Sound Perception to Identify Candidate Frequency for Wireless Networking

Wireless technology has been introduced and growing since early twentieth century, but there are still environments the current technologies find it difficult to penetrate. The dense jungle terrain, for example, pose a huge challenge for the 0.12 m wavelength of the Wi-Fi signals, but the FM radio frequency signals at a wavelength of 3 m function a lot better. This paper studies the possibility of using a very low frequency, down to the range of audible frequencies to try and identify the frequency band that can be used, ubiquitously and unobtrusively. Sound can be considered as a ubiquitous signal due to obvious reasons and the search is to find the unobtrusive frequency band that can be a candidate frequency for data carrier signals. The paper is presented in two sections, the first section does a geographically and age neutral survey to identify the unobtrusive signal and second section analyses the noise profiles in these frequency bands.

Distributed Energy Generation—Is It the Way of the Future?

This research topic consists of the advantages of distributed energy generation, the different methods used to generate the energy (e.g. photovoltaic panel and wind turbine), how to store the energy (e.g. battery) and how it will be achieved. Other topics that are discussed are cost factor and performance of each method. Furthermore, differences between the traditional transmission and distribution grid are compared. Lastly, the implementation of a smart grid is discussed to show that “green power” can provide a significant environmental benefit.

A study of open ports as security vulnerabilities in common user computers

The unprecedented growth in information technology and information explosion, with more and more data in electronic forms has put the computer into the hands of users with very little technical knowledge. The fact that the systems are not inherently immune, and that open up a number of vulnerabilities, leading to potential attacks, and the most prominent is in the form of open ports. This paper attempts to do a survey on common user computing devices including start devices to discover open ports and thereby explore vulnerabilities that can lead to potential attack targets.

Experimental comparison of uninformed and heuristic AI algorithms for N puzzle and 8 queen puzzle solution

This paper compares the performance of popular AI techniques, namely the Breadth First Search, Depth First Search, A* Search, Greedy Best First Search and the Hill Climbing Search in approaching the solution of a N-Puzzle of size 8, on a 3x3 matrix board and for the solution of the classic 8 Queen Puzzle. It looks at the complexity of each algorithm as it tries to approaches the solution in order to evaluate the operation of each technique and identify the better functioning one in various cases. The N Puzzle and the 8 Queen is used as the test scenario. An application was created to implement each of the algorithms to extract results for each of the cases. The paper also depicts the extent each algorithm goes through while processing the solution and hence helps to clarify the specific cases in which a technique may be preferred over another.