A speculative parallel simulated annealing algorithm based on Apache Spark

Abstract

Simulated annealing (SA) is an effective method for solving unconstrained optimization problems and has been widely used in machine learning and neural network. Nowadays, in order to optimize complex problems with big data, the SA algorithm has been implemented on big data platform and obtains a certain speedup. However, the efficiency for such implementation is still limited because the conventional SA algorithm still runs with low parallelism on new platforms and the computing resource cannot be fully utilized. For these problems, this paper raised a speculative parallel SA algorithm based on Apache Spark to expand the algorithm's parallelism and enhance its efficiency. In this paper, first, the inner dependencies, which stop conventional algorithm, run in parallel, are analyzed. Then, based on the analysis, the Software Thread-Level Speculation technique is employed to help the conventional algorithm overcome the dependencies and make it run concurrently. Finally, a new parallel SA algorithm with speculation mechanism is proposed and implemented on Apache Spark. The experiments show that, for big data problems, the proposed algorithm could achieve an optimal parallelism when comparing the traditional algorithm without speculation on Apache Spark. Moreover, the execution efficiency of simulated annealing process can be markedly enhanced by the proposed algorithm. KEYWORDSApache Spark, parallel computing, simulated annealing, thread level speculation INTRODUCTIONSimulated annealing (SA) is a probabilistic technique for approximating the global optimum of a given function. Specifically, it is a metaheuristic to approximate global optimization in a large search space. 1 However, with the data growing larger, traditional SA algorithm is barely to fulfill the present speed request of data processing. This is because, for one thing, the traditional method is inefficient and poorly paralleled, for another, the computing power the present platforms could afford is not enough to support large scale processing.To solve this problem, parallel programming is frequently used in accelerating the conventional SA algorithm. Researchers have successfully implemented parallelized SA algorithm on multicore platforms and achieved certain speedup. Moreover, a good approach to process large-scale data is to make use of enormous computing power offered by modern distributed computing platforms (or called big data platforms) like ApacheHadoop 2 or Apache Spark. 3 These popular platforms adopt MapReduce, which is a specialization of the split-apply-combine strategy for data analysis, as their programming model. A standard MapReduce program is composed by pairs of Map operations (whose job is to sorting or filtering data) and Reduce operations (whose job is to do summary of the result by Map operations), and a high parallelism of a MapReduce model is obtained by marshalling the operation pairs and performing them in distributed servers in parallel. The platforms that adopt MapReduce model often split the input, turn the large scale prob...

A Storage Optimization Model for Cloud Servers in Integrated Communication, Sensing, and Computation System

“…In the practical production environment, a trimendous amount of multisource heterogeneous data is generated in the custom manufacturing process of complex heavy equipment, and accordingly, its cloud service side will also generate ten million-level system logs. In order to process massive logs offline in real time, we use Apache Spark as the implementation platform of the proposed model [23]. By transforming system logs into elastic datasets using the machine learning tool called MLlib in Apache Spark, we can quickly analyze stored files and provide timely feedback to clients for deletion.…”

Water Allocation Optimization and Environmental Planning with Simulated Annealing Algorithms

“…When the development of each district is similar, it is allocated to each district according to the proportion of water demand; if the development levels of the two places are inconsistent, priority should be given to efficiency. Appropriately tilt towards high-efficiency industries so that the guarantee rate of high-efficiency water industries is higher than that of low-efficiency water industries [21].…”

An Annealing Accelerator for Ising Spin Systems Based on In‐Memory Complementary 2D FETs

“…SA provides a simple framework which can be implemented on systems with arbitrary energy landscapes, and it statistically guarantees an optimal solution. SA has hence been employed to solve optimization problems in a wide variety of domains such as circuit design, [ 5 ] data analysis, [ 6 ] imaging, [ 7 ] neural networks, [ 8 ] geophysics, [ 9 ] finance, [ 10 ] and the Ising model of magnetism. [ 11 ]…”

“…SA provides a simple framework which can be implemented on systems with arbitrary energy landscapes, and it statistically guarantees an optimal solution. SA has hence been employed to solve optimization problems in a wide variety of domains such as circuit design, [5] data analysis, [6] imaging, [7] neural networks, [8] geophysics, [9] finance, [10] and the Ising model of magnetism. [11] SA draws inspiration from physical annealing, in which a material is heated above its recrystallization temperature to allow atoms to rearrange and is then slowly cooled down to improve its crystallinity and reach a low energy state.…”