Time filter

Source Type

Sreelaja N.K.,Sri Krishna College of Engineering And Technology | Vijayalakshmi Pai G.A.,PSG College of Technology
Applied Soft Computing Journal | Year: 2012

Encryption of binary images is essential since it is vulnerable to eavesdropping in wired and wireless networks. The security of data becomes important since the communications over open network occur frequently. This paper focuses on encryption of binary image using a stream cipher method. In this paper we propose an Ant Colony Optimization (ACO) based approach of generating keys for encryption. The binary image is represented in a text form and encrypted using a stream cipher method. A novel technique termed Ant Colony Optimization Key Generation Binary Image Encryption (AKGBE) algorithm employs a character code table for encoding the keys and the plain text representing the binary image. The main advantage of this approach is that it reduces the number of keys to be stored and distributed. Experimental results demonstrating AKGBE's encrypting binary images of different sizes and the comparison of its performance with other stream cipher methods are presented. © 2012 Elsevier B.V. All rights reserved.

Jawahar N.,Thiagarajar College of Engineering | Balaji N.,Sri Krishna College of Engineering And Technology
Applied Soft Computing Journal | Year: 2012

This paper proposes a genetic algorithm (GA) based heuristic to the multi-period fixed charge distribution problem associated with backorder and inventories. The objective is to determine the size of the shipments, backorder and inventories at each period, so that, the total cost incurred during the entire period towards transportation, backorder and inventories is minimum. The model is formulated as pure integer nonlinear programming and 0-1 mixed integer linear programming problems, and proposes a GA based heuristic to provide solution to the above problem. The proposed GA based heuristic is evaluated by comparing their solutions with lower bound, LINGO solver and approximate solutions. The comparisons reveal that the GA generates better solutions than the approximate solutions, and is capable of providing solutions equal to LINGO solutions and closer to the lower bound value of the problems. © 2011 Elsevier B.V. All rights reserved.

Sreeja N.K.,Sri Krishna College of Engineering And Technology | Sankar A.,PSG College of Technology
Swarm and Evolutionary Computation | Year: 2016

Most data mining algorithms aim at discovering customer models and classification of customer profiles. Application of these data mining techniques to industrial problems such as customer relationship management helps in classification of customers with respect to their status. The mined information does not suggest any action that would result in reclassification of customer profile. Such actions would be useful to maximize the objective function, for instance, the net profit or minimizing the cost. These actions provide hints to a business user regarding the attributes that have to be changed to reclassify the customers from an undesirable class (e.g. disloyal) to the desired class (e.g. loyal). This paper proposes a novel algorithm called Hierarchical Heterogeneous Ant Colony Optimization based Action Rule Mining (HHACOARM) algorithm to generate action rules. The algorithm has been developed considering the resource constraints. The algorithm has ant agents at different levels in the hierarchy to identify the flexible attributes whose values need to be changed to mine action rules. The advantage of HHACOARM algorithm is that it generates optimal number of minimal cost action rules. HHACOARM algorithm does not generate invalid rules. Also, the computational complexity of HHACOARM algorithm is less compared to the existing action rule mining methods. © 2016 Elsevier B.V.

Sreeja N.K.,Sri Krishna College of Engineering And Technology | Sankar A.,PSG College of Technology
Applied Soft Computing Journal | Year: 2015

Abstract Classification is a method of accurately predicting the target class for an unlabelled sample by learning from instances described by a set of attributes and a class label. Instance based classifiers are attractive due to their simplicity and performance. However, many of these are susceptible to noise and become unsuitable for real world problems. This paper proposes a novel instance based classification algorithm called Pattern Matching based Classification (PMC). The underlying principle of PMC is that it classifies unlabelled samples by matching for patterns in the training dataset. The advantage of PMC in comparison with other instance based methods is its simple classification procedure together with high performance. To improve the classification accuracy of PMC, an Ant Colony Optimization based Feature Selection algorithm based on the idea of PMC has been proposed. The classifier is evaluated on 35 datasets. Experimental results demonstrate that PMC is competent with many instance based classifiers. The results are also validated using nonparametric statistical tests. Also, the evaluation time of PMC is less when compared to the gravitation based methods used for classification. © 2015 Elsevier B.V. All rights reserved.

Kotteeswaran R.,St. Josephs College | Sivakumar L.,Sri Krishna College of Engineering And Technology
Journal of Process Control | Year: 2014

Coal gasifier, an essential part of Integrated Gasification Combined Cycle (IGCC) converts coal into synthesis gas (syngas or producer gas) under certain pressure and temperature. The quality of syngas is highly influenced by quality of coal (calorific value) and hence greatly affects the power generation. Gasifier control seems to be highly difficult since it involves many variables and inherent nonlinearity. The baseline PI controller provided with ALSTOM benchmark challenge II (benchmark model of coal gasifier) fails to satisfy the constraints at 0% load for sinusoidal pressure disturbance and coal quality variations (±18%). This paper evaluates the tuning parameters of ALSTOM benchmark challenge II using Multi-Objective Particle Swarm Optimisation (MOPSO) algorithm. Robustness of the optimal PI controller is tested under sinusoidal and step pressure disturbance tests at 100%, 50% and 0% load conditions with decreased and increased coal quality variations. Test results show that the optimal PI controller meets all the constraints comfortably at all load conditions and provides better results for coal quality variations. © 2013 Elsevier Ltd. All rights reserved.

Discover hidden collaborations