An efficient intrusion detection mechanism based on particle swarm optimization and KNN

In this paper an efficient intrusion detection mechanism based on particle swarm optimization and KNN has been presented. In our approach experimentation has been performed for the intrusion detection considering NSL-KDD dataset. Then the selected weights have been added directly to the final classification which has been received safely. Then the remaining selected weights have been added for the classification. These nodes are originally safe but received unsafe. It has been input for the classification process. KNN has been used for the classification of the initial features and the content features. The remaining features have been transferred to the particle swarm optimization. PSO has been used for the classification of the traffic and host features. It has been classified based on 50% threshold value. The results show that by using our approach the average classification accuracy is approximately 98%. The attack considered here are Denial of Service (DoS), User to Root (U2R), Remote to User /Login (R2L) and Probe.

An efficient image cryptography mechanism based on the hybridization of standard encryption algorithms

In our approach the combination of RC4 and Blowfish algorithms with chaos mapping (RC4BC) has been presented for the image encryption. In the first phase image dataset has been considered. Then weight values have been calculated based on the image pixels. It has been calculated based on the attribute pixel, corelated pixels and the analytical factors of the edges. Then substitution has been performed through RC4 encryption mechanism. The substitution has been performed based on 8 sub blocks for the first time. After that different random substation has been applied based on the reallocation of bytes up to 16 rounds. Final substitution has been performed by the blowfish and chaos mapping. By this method the pixels of the images are rotated and shuffled with the XOR operation along with 16-byte reshuffling in each iteration. A total of 50 rounds have been considered for the shuffling, rotation and mapping. Then for comparative analysis peak signal to noise ratio (PSNR) and mean square error (MSE) have been calculated. Our results show that it is efficient in terms of MSE and PSNR values.

An analytical survey on the role of machine learning algorithms in case of intrusion detection

In this paper an analytical survey on the role of machine learning algorithms in case of intrusion detection has been presented and discussed. This paper shows the analytical aspects in the development of efficient intrusion detection system (IDS). The related study for the development of this system has been presented in terms of computational methods. The discussed methods are data mining, artificial intelligence and machine learning. It has been discussed along with the attack parameters and attack types. This paper also elaborates the impact of different attack and handling mechanism based on the previous papers.

An efficient data security mechanism with data sharing and authentication

In this paper an efficient security mechanism has been adopted for the cloud computing environment. It also provides an extendibility of cloud computing environment with big data and Internet of Things. AES-256 and RC6 with two round key generation have been applied for data and application security. Three-way security mechanism has been adopted and implemented. It is user to user (U to U) for data sharing and inter cloud communication. Then user to cloud (U to C) for data security management for application level hierarchy of cloud. Finally, cloud to user (C to U) for the cloud data protection. The security analysis has been tested with different iterations and rounds and it is found to be satisfactory.

An efficient clustering mechanism in big data framework for data preprocessing and management

An efficient data handling mechanism has been applied based on epoch-based k-means associated fuzzy clustering (EKFC). In the first phase weights have been assigned to individual data segment presented based on the classification key metrics. It has been assigned automatically. Then weight preprocessing has been done in such manner to prune the unwanted weights. It has been pruned in such way to filter the weights which are not scalable. Then epoch-based k-means associated fuzzy clustering (EKFC) approach has been applied for data arrangement. First different epochs have been considered for the calculation of initial seeds values. These seeds have been considered after considering 100 epochs. After 100 epochs seeds have been determined. These seeds values have been used as the initial centroid for the k-means clustering. After the complete validation similar clusters from the two clustering approaches have been considered. In the next phase operational clustering has been performed. In the final phase threshold ranking has been performed. It has been performed for the final classification based on the above clusters. It will arrange in the order of threshold values. It will be used for the determination of the priority of the task in the big data environment. The results are found to be prominent in terms of classification accuracy.

An efficient logistic regression and ant colony optimization-based object-oriented quality prediction

In this paper an efficient logistic regression and ant colony optimization-based object-oriented quality prediction has been presented. The dataset has been considered based on the object-oriented codes. The code considered here are completely equipped with object-oriented features for the complete analysis. The major parameters considered for the experimentation are class, object, inheritance and dynamic behavior. Class and objects have been considered for the class labels and memory requirements checking with the proper stack call and constructor invocation. Inheritance has been considered for the code reusability testing. Dynamic behavior has been tested for the runtime allocation. Then clustering algorithm is used for the parameter preprocessing and grouping based on the parameters. For data filtration chi-square testing has been applied. Then logistic regression and ant colony optimization (LR-ACO) have been considered for the final classification. Then F-Measure, Odd Ratio and Power have been considered for the analysis of the classification based on LR-ACO. The result after LR-ACO shows better accuracy as comparison to the previous methods.

Object oriented quality prediction through artificial intelligence and machine learning: a survey

Software quality estimation is an important aspect as it eliminates design and code defects. Object- oriented quality metrics prediction can help in the estimation of software quality of any defects and the chances of errors. In this paper a survey and the case analytics have been presented for the object-oriented quality prediction. It shows the analytical and experimental aspects of previous methodologies. This survey also elaborates different object-oriented parameters which is useful for the same problem. It also elaborates the problem aspects as well the limitations for the future directions. Machine learning and artificial intelligence methods have been considered mostly for this survey. The parameters considered are inheritance, dynamic behavior, encapsulation, objects etc.

Survey and analysis of cloud data management and security

This paper explores different security aspects in cloud computing environment. It includes data sharing mechanism, inter cloud communication, data breaches, data control, user-cloud relationship along with the cloud data management with standard security algorithms. It also covers the related reviews and analytical analysis on the traditional approaches for the gap identification. So, a short meta-analysis has been presented based on the method discussed along with the advantages and challenges found. It also explores the future prospective where there is the need of exploration and research.