Search results for ""author thomas s clary""

Volume 19 first describes the principles, materials and processes in 2.5D printing and post-press enhancement based on the authors' research. This compilation goes on to discuss Hadoop, a pioneering open source framework that has revolutionized the big data world due to its ability to process vast amounts of unstructured and semi-structured data by distributing a massive amount of data across clusters. A new tool that uses general and query-based summarization to analyze students' log files at the Academic Center for Tutoring at Columbus State University, a small state university in Georgia, USA, is introduced. The concept of a Vehicle Accident Prediction System is presented using accident datasets based on actual vehicle accident data extracted from open-source datasets. The current problem of rapid data increase throughout the year is discussed, along with the available options for gaining useful insight from World Wide Web created data. A review of different methods for Audio-Visual Speech Recognition using Random Forest is presented, and a strategy based on the combination of Wavelet multiresolution analysis and Random Forest is proposed. The authors review six recent applications of autoencoders in cryo-EM and cryo-ET data analysis, discussing the autoencoders' strength and weaknesses to provide potential directions for future research. In closing, an algorithm is presented which uses ensembles of metamodels and classifiers to predict which candidate designs for computer simulations are expected to cause failures and to divert the search accordingly.

Read more less

Volume 18 first discusses intermittently connected networks, a form of the delay tolerant network, where there never exists a complete end-to-end path between two nodes wishing to communicate. The authors provide an overview of the current algorithms (both analytic and iterative) for the generation of tomographic images, including comparisons of the basic characteristics of image quality. An intelligent scheduling framework is proposed to alleviate the problems encountered in modern job scheduling. The challenges of high performance computing scheduling and state-of-art scheduling methods to overcome these challenges are investigated. The state of art developments in quantum cryptography are presented, as well as the fundamental laws of quantum physics that quantum key distribution is based on. Subsequently, three main protocols of key distribution are introduced which enable users to share a secret between them. Three iterative methods for dose reconstruction in protoacoustics are discussed, including: gradient-descent algorithm, SART and MLEM, which are widely used in other medical imaging modalities such as PET, SPECT and CT. Additionally, the authors evaluate the behaviour of the concurrent execution of tasks in different implementations by Java virtual machines. The performance measurements were analysed by rigorous statistical techniques: analysis of Variance and HSD comparison test of Tukey averages. A mathematical model for finding the distribution of threads in local pools that minimises the makespan in integration processes is also presented. Makespan is a metric that calculates the average time a message takes to be executed by an integration process. In closing, varied knowledge sources for aspect clustering are explored a new method, OpCluster-PT, is proposed. To support this research, annotated corpora and some related resources that are publicly available to the interested reader have been included.

Read more less

This compilation opens with a review of the applications of several anomaly-based methods under the computational intelligence umbrella for the detection of DDoS attacks. Following this, a study is carried out to reveal the effects of a device developed to enable visually-impaired people to read any document in the Braille alphabet. The performance of the Artificial Bee Colony (ABC) algorithm on CEC2010 benchmark problems is also studied, with the goal of increasing the performance of the algorithm changes presented in large-scale optimization problems. Later, the opportunities and limitations of present waste management techniques are highlighted, and some future research proposals are discussed. The authors provide an overview of the field of motion capture focusing on methods, systems, and applications. More information about the motion processing and motion reconstruction technologies applied in the most prevalent optical and inertial systems is presented. In closing, a partially-manual method for using images to measure body poses is presented and discussed.

Read more less