Madhabananda Das

A Multi-objective Cat Swarm Optimization Algorithm for Workflow Scheduling in Cloud Computing Environment

As the world is progressing towards faster and more efficient computing techniques, cloud computing has emerged as an efficient and cheaper solution to such increasing and demanding requirements. Cloud computing is a computing model which facilitates not only the end-users but also organizational and other enterprise users with high availability of resources on demand basis. This involves the use of scientific workflows that require large amount of data processing, which can be costly and time-consuming if not properly scheduled in cloud environment. Various scheduling strategies have been developed, which include swarm-based optimization approaches as well. Due to the presence of multiple and conflicting requirements of users, multi-objective optimization techniques have become popular for workflow scheduling. This paper deals with cat swarm-based multi-objective optimization approach to schedule workflows in a cloud computing environment. The objectives considered are minimization of cost, makespan and CPU idle time. Proposed technique gives improved performance, compared with multi-objective particle swarm optimization (MOPSO) technique.

A New Approach to Associative Classification Based on Binary Multi-objective Particle Swarm Optimization

Associative classification rule mining (ACRM) methods operate by association rule mining (ARM) to obtain classification rules from a previously classified data. In ACRM, classifiers are designed through two phases: rule extraction and rule selection. In this paper, the ACRM problem is treated as a multi-objective problem rather than a single objective one. As the problem is a discrete combinatorial optimization problem, it was necessary to develop a binary multi-objective particle swarm optimization (BMOPSO) to optimize the measure like coverage and confidence of association rule mining (ARM) to extract classification rules in rule extraction phase. In rule selection phase, a small number of rules are targeted from the extracted rules by BMOPSO to design an accurate and compact classifier which can maximize the accuracy of the rule sets and minimize their complexity simultaneously. Experiments are conducted on some of the University of California, Irvine (UCI) repository datasets. The comparative result of the proposed method with other standard classifiers confirms that the new proposed approach can be a suitable method for classification.

A Review of Software Cost Estimation in Agile Software Development Using Soft Computing Techniques

For a successful software project, accurate prediction of its overall effort and cost estimation is a very much essential task. Software projects have evolved through a number of development models over the last few decades. Hence, to cover an accurate measurement of the effort and cost for different software projects based on different development models having new and innovative phases of software development, is a crucial task to be done. An accurate prediction always leads to a successful software project within the budget with no delay, but any percentage of misconduct in the overall effort and cost estimate may lead to a project failure in terms of delivery time, budget or features. Software industries have adopted various development models based on the project requirements and organizations capabilities. Due to adaptability to changes in a software project, agile software development model has become a much successful and popular framework for development over the last decade. The customer is involved as an active participant in the development using an agile framework. Hence, changes can occur at any phase of development and they can be dynamic in nature. That is why an accurate prediction of effort and cost of such projects is a crucial task to be done as the complexity of overall development structure is increased with the time. Soft computing techniques have proven that they are one of the best problem solving techniques in such scenarios. Such techniques are more flexible and presence of bio-intelligence increases their accuracy. Genetic Algorithm (GA), Particle Swarm Optimization (PSO), Artificial Neural Network (ANN), Fuzzy Inference Systems (FIS), etc. are applied successfully for estimation of cost and effort of agile based software projects. This paper deals with such soft computing techniques and provides a detailed and analytical overview of such methods. It also provides the future scope and possibilities to explore such techniques on the basis of survey provided by this paper.

Implementation of an optimized classification model for prediction of hypothyroid disease risks

Data mining based classification is one of the important role in the field of healthcare. Diagnosis of health conditions is a very important and challenging task in field of medical science. There are various types of diseases are diagnosis in medical science. Thyroid disease is one of critical diseases that is very serious problem and affected the health of human being. Thyroid decease classification is one of the important problems in medical science because it is directly related to health condition of human body, this type of disease can be solve by proper identify and carefully treatment. This paper focuses on the survey of diagnosis of thyroid. There are various authors have worked in the field of thyroid diseases classification and give the classification accuracy with robust model. This research is also focus on the various techniques that is applied for classification of thyroid data.

Medical Image Thresholding Using Particle Swarm Optimization

Image processing has been serving as one major part of medical science since 1980s as automation of image analysis offers better results in efficient time period to help specialists in diagnosis and eradication of diseases. Most frequently, medical fields face different cases of detecting tumors, kidney stones, fractures in bones, etc. through various images such as ultrasound images and X-ray images. But it is very difficult for identification of some particular structure in some medical images. Hence, such images need more improvement in terms of noise reduction and segmentation. Image thresholding is a kind of segmentation process which partitions the image into different objects. Particle swarm optimization (PSO) is one bio-inspired optimization technique which gets one optimized threshold value for image thresholding in this paper using proper fitness function.

Detection and Reduction of Impulse Noise in RGB Color Image Using Fuzzy Technique

A concept of impulse noise reduction method for an RGB color image with a fuzzy detection phase is introduced and a fuzzy de-noising procedure is used to filter the color image. In this paper, each color component is correlated to the other two corresponding color components to overcome the color disorder on edge and texture pixel. Here the filtering technique is only applied to noisy pixel, detected by fuzzy technique, while preserving the color and edge sharpness. Experimental results show that the proposed method provides noteworthy improvement on other non-fuzzy and fuzzy filters.

Architectural layout planning using genetic algorithms

This paper presents a model which generates architectural layout for a single flat having regular shaped spaces; Bedroom, Bathroom, Kitchen, Balcony, Living and Dining Room. Using constraints at two levels; Topological (Adjacency, Compactness, Vaastu, and Open and Closed face constraints) and Dimensional (Length to Width ratio constraint), Genetic Algorithms have been used to generate the topological arrangement of spaces in the layout and further if required, feasibility have been dimensionally analyzed. Further easy evacuation form the selected layout in case of adversity has been proposed using Dijkstras Algorithm. Later the proposed model has been tested for efficiency using various test cases. This paper also presents a classification and categorization of various problems of space planning.

Estimation of Incomplete Data in Mixed Dataset

This paper puts forward a fresh approach which is a modification of original fuzzy kNN for dealing with categorical missing values in categorical and mixed attribute datasets. We have removed the irrelevant missing samples through list-wise deletion. Then, rest of the missing samples is estimated using kernel-based fuzzy kNN technique and partial distance strategy. We have calculated the errors at different percentage of missing values. Results highlight that mixture kernel gives minimum average of MAE, MAPE and RMSE at different missing percentage when implemented on lenses, SPECT heart and abalone dataset.

Learning to detect SQLIA using node centrality with feature selection

Web applications hosted on the Internet are naturally exposed to a variety of attacks and constantly probed by hackers for vulnerabilities. SQL Injection Attack (SQLIA) has been a major security threat on web applications since over 15 years. Detecting SQLIA at runtime is a challenging problem because of extreme heterogeneity of the attack vectors. This paper explores application of node centrality metrics to train a Support Vector Machine (SVM) for identifying malicious queries containing SQL injection attacks. The WHERE clause portion of SQL queries are first normalized into a sequence of tokens and then modeled as interaction networks, from which centrality of the nodes are computed. After applying feature selection by information gain method, the centrality scores of high ranking nodes are used to train the SVM classifier. We experiment with four centrality measures popularly used in Social Network Analysis (SNA). The results on five sample web applications built with PHP/MySQL show that this technique can effectively detect SQLIA with minimal performance overhead. Designed for the database firewall layer, the approach can protect multiple websites on a shared server, which is another advantage.

Predicting Users’ Preferences for Movie Recommender System Using Restricted Boltzmann Machine

Recommender system is one of the most important crucial parts for e-commerce domains, enabling them to produce correct recommendations to individual users. Collaborative filtering is considered as the successful technique for recommender system that takes rating scores to find most similar users/items for recommending items. In this work, in order to exploit user rating information, a model has been developed that uses Restricted Boltzmann Machine (RBM) to learn deeply and predict the ratings or preferences which are missed. The experiment is done on MovieLens benchmark dataset that compares with Pearson correlation and average prediction-type algorithms. Experimental result exhibits the performance of RBM to predict users’ preferences.