Arobinda Gupta

Ashutosh Baheti, Arobinda Gupta

A belt region is said to be k-barrier covered by a set of sensors if all paths crossing the width of the belt region intersect the sensing regions of at least k sensors. Barrier coverage can be achieved from a random initial deployment of mobile sensors by suitably relocating the sensors to form a barrier. Reducing the movement of the sensors is important in such scenarios due to the energy constraints of sensor devices. In this paper, we propose a centralized algorithm which achieves 1-barrier coverage by forming a non-linear barrier from a random initial deployment of sensors in a belt. The algorithm uses a novel idea of physical behavior of chains along with the concept of virtual force. Formation of non-linear barrier reduces the movement of the sensors needed as compared to linear barriers. Detailed simulation results are presented to show that the proposed algorithm achieves barrier coverage with less movement of sensors compared to other existing algorithms in the literature.

Anirban Santara, Debapriya Maji, DP Tejas et al.

Deep neural networks are capable of modelling highly non-linear functions by capturing different levels of abstraction of data hierarchically. While training deep networks, first the system is initialized near a good optimum by greedy layer-wise unsupervised pre-training. However, with burgeoning data and increasing dimensions of the architecture, the time complexity of this approach becomes enormous. Also, greedy pre-training of the layers often turns detrimental by over-training a layer causing it to lose harmony with the rest of the network. In this paper a synchronized parallel algorithm for pre-training deep networks on multi-core machines has been proposed. Different layers are trained by parallel threads running on different cores with regular synchronization. Thus the pre-training process becomes faster and chances of over-training are reduced. This is experimentally validated using a stacked autoencoder for dimensionality reduction of MNIST handwritten digit database. The proposed algorithm achieved 26\% speed-up compared to greedy layer-wise pre-training for achieving the same reconstruction accuracy substantiating its potential as an alternative.