A5010537832- Parallel Computing and Optimization Techniques
- Data Mining Algorithms and Applications
- Distributed and Parallel Computing Systems
- Distributed systems and fault tolerance
- Data Stream Mining Techniques
- Advanced Multi-Objective Optimization Algorithms
- Metaheuristic Optimization Algorithms Research
- Machine Learning and Data Classification
- Advanced Bandit Algorithms Research
- Vehicle Routing Optimization Methods
- Logic, programming, and type systems
- Data Management and Algorithms
- Evolutionary Algorithms and Applications
- Eating Disorders and Behaviors
- Rough Sets and Fuzzy Logic
- Patient Satisfaction in Healthcare
- Interconnection Networks and Systems
- Embedded Systems Design Techniques
- Software Engineering Research
- Behavioral Health and Interventions
- Obesity, Physical Activity, Diet
- Advanced Data Storage Technologies
- Healthcare Policy and Management
- Real-Time Systems Scheduling
- Cloud Computing and Resource Management
Manchester Metropolitan University
2016-2019
Medical Solutions
2019
University of Manchester
2008-2015
Intel (United States)
2010
Effective Big Data Mining requires scalable and efficient solutions that are also accessible to users of all levels expertise. Despite this, many current efforts provide effective knowledge extraction via large-scale tools focus more on performance than use tuning which complex problems even for experts. Weka is a popular comprehensive workbench with well-known intuitive interface, nonetheless it supports only sequential single-node execution. Hence, the size datasets processing tasks can...
Fundamental nano-patterns are simple, static, binary properties of Java methods, such as ObjectCreator and Recursive. We present a provisional catalogue 17 nano-patterns. report statistical information theoretic metrics to show the frequency nano-pattern occurrence in large corpus open-source projects. proceed give two example case studies that demonstrate potential applications for The first study involves quantitative comparison popular benchmarking suites, terms their relative...
Thread-Level Speculation (TLS) overcomes limitations intrinsic with conservative compile-time auto-parallelizing tools by extracting parallel threads optimistically and only ensuring absence of data dependence violations at runtime. A significant barrier for adopting TLS (implemented in software) is the overheads associated maintaining speculative state. Based on previous limit studies, we observe that future multicore systems will likely have more cores idle than those which traditional...
Thread-Level Speculation (TLS) facilitates the extraction of parallel threads from sequential applications. Most prior work has focused on developing compiler and architecture for this execution paradigm. Such studies often narrowly concentrated a specific design point. On other hand, have attempted to assess how well TLS performs if some architectural/ constraint is relaxed. Unfortunately, such previous failed truly performance potential, because they been bound ignored one or another...
Hadoop provides a scalable solution on traditional cluster-based Big Data platforms but imposes performance overheads due to only supporting on-disk data. Analytic algorithms usually require multiple iterations over dataset and thus, multiple, slow, disk accesses. In contrast, modern clusters possess increasing amounts of main memory that can provide benefits by efficiently using caching mechanisms. Apache Spark is an innovative distributed computing framework supports in-memory...
Current parallelizing compilers can tackle applications exercising regular access patterns on arrays or affine indices, where data dependencies be expressed in a linear form. Unfortunately, there are cases that independence between statements of code cannot guaranteed and thus the compiler conservatively produces sequential code. Programs involve extensive pointer use, irregular patterns, loops with unknown number iterations examples such cases. This limits extraction parallelism rarely...
Ant Colony Optimization (ACO) is a nature-inspired optimization metaheuristic which has been successfully applied to wide range of different problems. However, significant limiting factor in terms its scalability memory complexity; many problems, the pheromone matrix encodes trails left by ants grows quadratically with instance size. For very large instances, this requirement factor, making ACO an impractical technique. In paper we propose restricted variant linear complexity, stores values...
Ant Colony Optimization (ACO) is a well-established nature-inspired heuristic, and parallel versions of the algorithm now exist to take advantage emerging high-performance computing processors. However, careful attention must be paid components such implementations if full benefit these platforms obtained. One component ACO next node selection, which presents unique challenges in setting. In this paper, we present new selection method for ACO, Vectorized Candidate Set Selection (VCSS),...
SUDA2 is a recursive search algorithm for minimal unique itemset detection. Such sets of items are formed via combinations non-obvious attributes enabling individual record identification. The nature allows work to be divided into non-overlapping tasks parallel execution. Earlier developed implementation on an SMP cluster, and this was found several orders magnitude faster than sequential SUDA2. However, if fixed-granularity scheduled naively in the order their generation, system load tends...
Substantial research has been carried out to explain the effects of economic variables on obesity, typically considering only a few factors at time, using parametric linear regression models. Recent studies have made significant contribution by examining affecting body weight Behavioral Risk Factor Surveillance System data with 27 state-level for period 20 years (1990-2010). As elsewhere, authors solely focus individual potential drivers obesity than critical interactions among drivers. We...
The advent of multi-core processors, particularly with projections that numbers cores will continue to increase, has focused attention on parallel programming. It is widely recognized current programming techniques, including those are used for scientific programming, not allow the easy formulation general purpose applications. An area which receiving interest use styles side-effect free. Previous work functional demonstrated potential this permit exploitation parallelism. Recent systems...