A Software Architecture Assisting Workflow Executions on Cloud Resources

Abstract

An enterprise providing services handled by means of workflows needs to monitor and control their execution, gather usage data, determine priorities, and properly use computing cloud-related resources. This paper proposes a software architecture that connects unaware services to components handling workflow monitoring and management concerns. Moreover, the provided components enhance dependability of services while letting developers focus only on the business logic

Can We Process 2D Images Using Artificial Bee Colony?

“…Finally, we can introduce better fitness function. Furthermore, we will try to build a classifier for large collections [4] of input images using fast methods [16], [28] and [13].…”

“…Moreover, CI powers the resolution of complicated differential and integral equations like in positioning queueing systems: [7], [26], [18], [25] and [24]. CI can be applied in the optimization of optimal osmotic parameters (see [31]), optimal clustering methods (see [6]), evaluation in games (see [21], [20]), simulation of iron cast and heat transfer (see [9] and [10]), benchmark tests [30] and cloud-computing workflow management [4]. As You see CI can provide many methods that are efficient in various operations.…”

A mathematical model for file fragment diffusion and a neural predictor to manage priority queues over BitTorrent

“…Such modified values are computed by our RNNs, and then predicted future values for T k (τ n+r ), S k (τ n+r ) and ρ k (τ n+r ) are sent to each node active as a peer. Each time a new file becomes shared on the P2P BitTorrent network, a new RNN is created and trained on a server (e.g., requested from a cloud system (Borowik et al, 2015;Napoli et al, 2016)), in order to provide predictions related to peer availability of the novel set of shared fragments. Values indicating the prediction are sent to the peers periodically, and allow peers to update their values of D k (τ ).…”

Benchmark Tests on Improved Merge for Big Data Processing

“…Paralleled to this, in the common computer systems we need to have optimal traffic on the server that is managing our business network [7], [8], [9], [10]. This positioning and optimization can be efficiently solved by computational intelligence methods [11], [12], [13], [14], [15]. Finally, in the data base system, we need methods that are crucial for fastest possible ordering, sorting, feature extraction, data discretization, data retrieval or reconstruction [16], [17], [18], [19].…”