Computer science and engineering

We present a new high-quality, single-subject atlas with sub-millimeter voxel resolution, high SNR, and excellent grey-white tissue contrast to resolve fine anatomical details. The atlas is labeled into two parcellation schemes: 1) the anatomical BCI-DNI atlas, which is manually labeled based on known morphological and anatomical features, and 2) the hybrid USCBrain atlas, which incorporates functional information to guide the sub-parcellation of cerebral cortex. In both cases, we provide consistent volumetric and cortical surface-based parcellation and labeling. The intended use of the atlas is as a reference template for structural coregistration and labeling of individual brains. A single-subject T1-weighted image was acquired at a resolution of 0.547mm × 0.547mm × 0.800mm five times and averaged. Images were processed by an expert neuroanatomist using semi-automated methods in BrainSuite to extract the brain, classify tissue-types, and render anatomical surfaces. Sixty-six corti...

Load balancing involves assigning tasks to each processor proportional to its performance and to minimize communication overhead. The assignment can be static-done at compile time, or it may be dynamic-done at run-time. Many load balancing polices achieve high system performance by increasing the utilization of CPU, memory, or a combination of CPU and memory . In this paper Modified Triangle Scheduling Scheme (MTSS) is proposed which modifies the Minimum Distance Scheduling (MDS). This scheme has been implemented on Linearly Extensible Triangle (LEΔ) and Linearly Extensible Tree (LET) which reduces the Load Imbalance Factor (LIF) and also the execution time of parallel tasks assigned to the processors.

MANET is a multi-cluster, multi-hop packet radio network that should adapt itself to the dynamically changing network configurations. This dynamic reconfiguration is governed by clusterheads, a major set of nodes that have enough resources to be capable of routing packets in the network. This paper proposes a load balancing heuristic to extend the network lifetime. It has two phases. Initial clustering is done for the entire network in start up phase that takes node's degree and energy as input parameters. The maintenance phase keeps the network topology updated through global re-clustering and local updation. Global reclustering is initiated when the network becomes significantly unbalanced i.e. if the variance of degree of the cluster heads in the network is greater than a pre-determined threshold. It is expected to get better result than the time driven algorithms, as implementation is still under progress.

In this paper, an efficient and modified load balancing algorithm is proposed. This paper shows the results of the proposed energy aware load balancing algorithm. Global reclustering is initiated when the network becomes significantly unbalanced i.e. if the variance of degree of the cluster heads in the network is greater than a pre-determined threshold. Degree of all cluster heads would be required at a single node to evaluate the average of degree of all cluster heads in the network and subsequently the variance among themselves, which is not possible in adhoc network because of its distributed nature. For this, a formula is derived in which variance can be updated at each hop i.e. variance of N nodes can be expressed in terms of variance of (N-1) nodes and average of (N-1) nodes. The results of simulation indicate that compared with the load balanced routing algorithm DEECA, the proposed algorithm increased the network lifetime, packet delivery ratio at different traffic.

One of the widely used models in Machine Learning is an Artificial Neural Network. Ingenuous and legitimate results are the utmost need of every model. These models work like a black box, and no factor as trust is considered. Machine Learning algorithms are stuck with the risk of relying on biases encoded in the data or algorithms used for predictions. Prejudice or Bias is the root cause of the poor performance of the model. A novel approach is proposed, named legitANN, which has used three levels of bias control and an AvgNew loss function to make the model legitimate and reliable. Median Absolute deviation is a statistical measure that can guarantee the learning of neural networks to achieve unbiased robustness. Median Absolute deviation used here to prove the effectiveness of the method legitANN. The work proved that feed forward neural network is better at approximating. But for uniformly dense sets or preprocessed data, a radial basis network may be a right choice. This paper presents legitANN a new model that work on bias control using radial basis network and achieve better performance in comparison to other Artificial Neural Network in terms of improved evaluation metrics values for link prediction.

We declare that this written submission represents our ideas in our own words and where others' ideas or words have been included. We have adequately cited and referenced the original sources. We also declare that we have adhered to all principles of academic honesty and integrity and have not misrepresented or fabricated or falsified any idea / data / fact / source in our submission. We understand that any violation of the above will be a cause for disciplinary action by the Institute and can also evoke penal action from the sources which have thus not been properly cited or from whom proper permission has not been taken when needed. Abstract Transmission of a packet from a source node to a destination node efficiently in a wireless network having finite number of nodes, necessitates the need to reduce the number of link layer transmissions. The problem of computing the path with minimum cost in terms of the expected number of link layer transmissions(including retransmissions) in wireless mesh networks has been solved [1]. A non-commutative routing metric, Expected number of Transmissions On a Path(ETOP), for wireless mesh networks has been proposed, based on the observation that the performance of a path depends not only on the number of links on the path and the quality of its links, but also on the relative positions of the links on the path. ETOP, which is a non-commutative function of the link success probabilities, accurately captures the expected number of link layer transmissions required for reliable end-to-end packet delivery.

Concolic testing tools can automatically generate test cases. However, there is a scope to increase the coverage achieved. Pair-wise testing is a prominent black-box testing method. We propose to combine ideas from these two techniques to enhance the code coverage. We have carried out experimental studies to evaluate the effectiveness of our approach. Our experimental results show that on an average our approach achieves 3.27%, 2.94%, and 6.22% improvements in condition coverage, decision coverage, and MC/DC respectively over a traditional concolic tester. Further, on an average of 4.36% improvement in mutation score is achieved.