PMCRI: a parallel modular classification rule induction framework

In a world where massive amounts of data are recorded on a large scale we need data mining technologies to gain knowledge from the data in a reasonable time. The Top Down Induction of Decision Trees (TDIDT) algorithm is a very widely used technology to predict the classification of newly recorded data. However alternative technologies have been derived that often produce better rules but do not scale well on large datasets. Such an alternative to TDIDT is the PrismTCS algorithm. PrismTCS performs particularly well on noisy data but does not scale well on large datasets. In this paper we introduce Prism and investigate its scaling behaviour. We describe how we improved the scalability of the serial version of Prism and investigate its limitations. We then describe our work to overcome these limitations by developing a framework to parallelise algorithms of the Prism family and similar algorithms. We also present the scale up results of a first prototype implementation.

Parallel induction of modular classification rules

The Distributed Rule Induction (DRI) project at the University of Portsmouth is concerned with distributed data mining algorithms for automatically generating rules of all kinds. In this paper we present a system architecture and its implementation for inducing modular classification rules in parallel in a local area network using a distributed blackboard system. We present initial results of a prototype implementation based on the Prism algorithm.

Non-uniform data distribution for communication-efficient parallel clustering

Global communicationrequirements andloadimbalanceof someparalleldataminingalgorithms arethe major obstacles to exploitthe computational power of large-scale systems. This work investigates how non-uniform data distributions can be exploited to remove the global communication requirement and to reduce the communication costin parallel data mining algorithms and, in particular, in the k-means algorithm for cluster analysis. In the straightforward parallel formulation of the k-means algorithm, data and computation loads are uniformly distributed over the processing nodes. This approach has excellent load balancing characteristics that may suggest it could scale up to large and extreme-scale parallel computing systems. However, at each iteration step the algorithm requires a global reduction operationwhichhinders thescalabilityoftheapproach.Thisworkstudiesadifferentparallelformulation of the algorithm where the requirement of global communication is removed, while maintaining the same deterministic nature ofthe centralised algorithm. The proposed approach exploits a non-uniform data distribution which can be either found in real-world distributed applications or can be induced by means ofmulti-dimensional binary searchtrees. The approachcanalso be extended to accommodate an approximation error which allows a further reduction ofthe communication costs. The effectiveness of the exact and approximate methods has been tested in a parallel computing system with 64 processors and in simulations with 1024 processing element

A parallel quantum histogram architecture

A parallel formulation of an algorithm for the histogram computation of n data items using an on-the-fly data decomposition and a novel quantum-like representation (QR) is developed. The QR transformation separates multiple data read operations from multiple bin update operations thereby making it easier to bind data items into their corresponding histogram bins. Under this model the steps required to compute the histogram is n/s + t steps, where s is a speedup factor and t is associated with pipeline latency. Here, we show that an overall speedup factor, s, is available for up to an eightfold acceleration. Our evaluation also shows that each one of these cells requires less area/time complexity compared to similar proposals found in the literature.

Efficient group communication for large-scale parallel clustering

Global communication requirements and load imbalance of some parallel data mining algorithms are the major obstacles to exploit the computational power of large-scale systems. This work investigates how non-uniform data distributions can be exploited to remove the global communication requirement and to reduce the communication cost in iterative parallel data mining algorithms. In particular, the analysis focuses on one of the most influential and popular data mining methods, the k-means algorithm for cluster analysis. The straightforward parallel formulation of the k-means algorithm requires a global reduction operation at each iteration step, which hinders its scalability. This work studies a different parallel formulation of the algorithm where the requirement of global communication can be relaxed while still providing the exact solution of the centralised k-means algorithm. The proposed approach exploits a non-uniform data distribution which can be either found in real world distributed applications or can be induced by means of multi-dimensional binary search trees. The approach can also be extended to accommodate an approximation error which allows a further reduction of the communication costs.

A parallel formulation for the simulation of a generic branch predictor

A parallel formulation for the simulation of a branch prediction algorithm is presented. This parallel formulation identifies independent tasks in the algorithm which can be executed concurrently. The parallel implementation is based on the multithreading model and two parallel programming platforms: pthreads and Cilk++. Improvement in execution performance by up to 7 times is observed for a generic 2-bit predictor in a 12-core multiprocessor system.

Bit-index sort: a fast non-comparison integer sorting algorithm for permutations

This paper describes a fast integer sorting algorithm, herein referred as Bit-index sort, which is a non-comparison sorting algorithm for partial per-mutations, with linear complexity order in execution time. Bit-index sort uses a bit-array to classify input sequences of distinct integers, and exploits built-in bit functions in C compilers supported by machine hardware to retrieve the ordered output sequence. Results show that Bit-index sort outperforms in execution time to quicksort and counting sort algorithms. A parallel approach for Bit-index sort using two simultaneous threads is included, which obtains speedups up to 1.6.

A Parallel Hardware Architecture for Scale and Rotation Invariant Feature Detection

This paper proposes a parallel hardware architecture for image feature detection based on the Scale Invariant Feature Transform algorithm and applied to the Simultaneous Localization And Mapping problem. The work also proposes specific hardware optimizations considered fundamental to embed such a robotic control system on-a-chip. The proposed architecture is completely stand-alone; it reads the input data directly from a CMOS image sensor and provides the results via a field-programmable gate array coupled to an embedded processor. The results may either be used directly in an on-chip application or accessed through an Ethernet connection. The system is able to detect features up to 30 frames per second (320 x 240 pixels) and has accuracy similar to a PC-based implementation. The achieved system performance is at least one order of magnitude better than a PC-based solution, a result achieved by investigating the impact of several hardware-orientated optimizations oil performance, area and accuracy.

Grid job scheduling using Route with Genetic Algorithm support

In 2006 the Route load balancing algorithm was proposed and compared to other techniques aiming at optimizing the process allocation in grid environments. This algorithm schedules tasks of parallel applications considering computer neighborhoods (where the distance is defined by the network latency). Route presents good results for large environments, although there are cases where neighbors do not have an enough computational capacity nor communication system capable of serving the application. In those situations the Route migrates tasks until they stabilize in a grid area with enough resources. This migration may take long time what reduces the overall performance. In order to improve such stabilization time, this paper proposes RouteGA (Route with Genetic Algorithm support) which considers historical information on parallel application behavior and also the computer capacities and load to optimize the scheduling. This information is extracted by using monitors and summarized in a knowledge base used to quantify the occupation of tasks. Afterwards, such information is used to parameterize a genetic algorithm responsible for optimizing the task allocation. Results confirm that RouteGA outperforms the load balancing carried out by the original Route, which had previously outperformed others scheduling algorithms from literature.

PARALLEL ALGORITHMS FOR MAXIMAL CLIQUES IN CIRCLE GRAPHS AND UNRESTRICTED DEPTH SEARCH

We present parallel algorithms on the BSP/CGM model, with p processors, to count and generate all the maximal cliques of a circle graph with n vertices and m edges. To count the number of all the maximal cliques, without actually generating them, our algorithm requires O(log p) communication rounds with O(nm/p) local computation time. We also present an algorithm to generate the first maximal clique in O(log p) communication rounds with O(nm/p) local computation, and to generate each one of the subsequent maximal cliques this algorithm requires O(log p) communication rounds with O(m/p) local computation. The maximal cliques generation algorithm is based on generating all maximal paths in a directed acyclic graph, and we present an algorithm for this problem that uses O(log p) communication rounds with O(m/p) local computation for each maximal path. We also show that the presented algorithms can be extended to the CREW PRAM model.

Branch and Bound Algorithm for Multiprocessor Scheduling

The multiprocessor task graph scheduling problem has been extensively studied asacademic optimization problem which occurs in optimizing the execution time of parallelalgorithm with parallel computer. The problem is already being known as one of the NPhardproblems. There are many good approaches made with many optimizing algorithmto find out the optimum solution for this problem with less computational time. One ofthem is branch and bound algorithm.In this paper, we propose a branch and bound algorithm for the multiprocessor schedulingproblem. We investigate the algorithm by comparing two different lower bounds withtheir computational costs and the size of the pruned tree.Several experiments are made with small set of problems and results are compared indifferent sections.

Tabu search algorithm for network synthesis

Large scale combinatorial problems such as the network expansion problem present an amazingly high number of alternative configurations with practically the same investment, but with substantially different structures (configurations obtained with different sets of circuit/transformer additions). The proposed parallel tabu search algorithm has shown to be effective in exploring this type of optimization landscape. The algorithm is a third generation tabu search procedure with several advanced features. This is the most comprehensive combinatorial optimization technique available for treating difficult problems such as the transmission expansion planning. The method includes features of a variety of other approaches such as heuristic search, simulated annealing and genetic algorithms. In all test cases studied there are new generation, load sites which can be connected to an existing main network: such connections may require more than one line, transformer addition, which makes the problem harder in the sense that more combinations have to be considered.

Parallel simulated annealing applied to long term transmission network expansion planning

The simulated annealing optimization technique has been successfully applied to a number of electrical engineering problems, including transmission system expansion planning. The method is general in the sense that it does not assume any particular property of the problem being solved, such as linearity or convexity. Moreover, it has the ability to provide solutions arbitrarily close to an optimum (i.e. it is asymptotically convergent) as the cooling process slows down. The drawback of the approach is the computational burden: finding optimal solutions may be extremely expensive in some cases. This paper presents a Parallel Simulated Annealing, PSA, algorithm for solving the long term transmission network expansion planning problem. A strategy that does not affect the basic convergence properties of the Sequential Simulated Annealing algorithm have been implementeded and tested. The paper investigates the conditions under which the parallel algorithm is most efficient. The parallel implementations have been tested on three example networks: a small 6-bus network, and two complex real-life networks. Excellent results are reported in the test section of the paper: in addition to reductions in computing times, the Parallel Simulated Annealing algorithm proposed in the paper has shown significant improvements in solution quality for the largest of the test networks.

Combining genetic algorithm and simulated annealing: A molecular geometry optimization study

We introduce a new hybrid approach to determine the ground state geometry of molecular systems. Firstly, we compared the ability of genetic algorithm (GA) and simulated annealing (SA) to find the lowest energy geometry of silicon clusters with six and 10 atoms. This comparison showed that GA exhibits fast initial convergence, but its performance deteriorates as it approaches the desired global extreme. Interestingly, SA showed a complementary convergence pattern, in addition to high accuracy. Our new procedure combines selected features from GA and SA to achieve weak dependence on initial parameters, parallel search strategy, fast convergence and high accuracy. This hybrid algorithm outperforms GA and SA by one order of magnitude for small silicon clusters (Si6 and Si10). Next, we applied the hybrid method to study the geometry of a 20-atom silicon cluster. It was able to find an original geometry, apparently lower in energy than those previously described in literature. In principle, our procedure can be applied successfully to any molecular system. © 1998 Elsevier Science B.V.

A 2-D delaunay refinement algorithm using an initial prerefinement from the boundary mesh

An improvement to the quality bidimensional Delaunay mesh generation algorithm, which combines the mesh refinement algorithms strategy of Ruppert and Shewchuk is proposed in this research. The developed technique uses diametral lenses criterion, introduced by L. P. Chew, with the purpose of eliminating the extremely obtuse triangles in the boundary mesh. This method splits the boundary segment and obtains an initial prerefinement, and thus reducing the number of necessary iterations to generate a high quality sequential triangulation. Moreover, it decreases the intensity of the communication and synchronization between subdomains in parallel mesh refinement. © 2008 IEEE.