Recent Patents on Computer Science - Volume 5, Issue 3, 2012

New experimental data on metabolites and enzymes induce high interest in metabolic modelling including metabolic flux calculations. Data analysis of metabolites, calculation of metabolic fluxes, pathways and their conditionspecific strengths is now possible by an advantageous combination of specific software. How can available software for metabolic modelling be improved from a computational point of view? A number of available and well established software solutions are first discussed individually. This includes information on software origin, capabilities, development and used methodology. Performance information is obtained for the compared software using provided example data sets. A feature based comparison shows limitations and advantages of the compared software for specific tasks in metabolic modelling. Often found limitations include third party software dependence, no comprehensive database management and no standard format for data input and output. Graphical visualization can be improved for complex data visualization and at the web based graphical interface. Other areas for development are platform independency, product line architecture, data standardization, open source movement and new methodologies along with the discussion of few of the patents related to Computational features.

In the recent years, the global burden of diabetes has become apparent - more than 366 million people worldwide are affected by this disease. At the same time, it is well-known that the treatment of diabetes is one of the most difficult therapies to manage. Thus, the key problem of diabetes therapy management is to predict the future blood glucose level of a diabetic patient from available current and past information about therapeutically valuable factors. The developed approaches and algorithms to accurate prediction respond to the strong need for better management of the disease. In this paper we retrace the development of the blood glucose predictors and make a review of the ones recently published in the patent literature as well as in journals on diabetes and data-mining. We mention strengths and weaknesses of the existing models and present a novel approach for addressing the problem that may seem superior to the state of the art methods. In particular, we show that the predictors based on the novel approach are able to perform up to 20 minutes ahead in the future at the level of clinical accuracy achieved by the commercially available devices for reading just the current blood glucose value.

This paper describes Patent Number U.S. 8,046,316 B2, titled “Cyclic Combinatorial Method and System”, issued by the US Patents and Trademarks Office on October 25, 2011. The patent is based on a combinatorial algorithm to solve cyclic optimization problems. First, the algorithm identifies cyclically distinct solutions of such problems by enumerating cyclically distinct combinations of the basic dual variables. In combinatorial terminology, this stage of the algorithm addresses the following question: given n cyclic objects, how many cyclically distinct combinations of m (m ≤ n) objects can be selected? Integrating the operations of partition and cyclic permutation, a procedure is developed for generating cyclically distinct selections (dual solutions). Subsequently, rules are described for recognizing the set of dominant solutions. Finally, primal-dual complementary slackness relationships are used to find the primal optimum solution. This patent has many potential applications in optimization problems with cyclic 0-1 matrices, such as network problems and cyclic workforce scheduling. The patent's applicability has been illustrated by efficiently solving several cyclic labor scheduling problems.

Self-organization of a collection of agents is a crucial concern in multi-agent systems that operate in open and dynamic environments. Most self-organizing mechanisms proposed in previous works tackle organization structure issues at design time. However, in open environments, agents must be able to adapt by achieving the most appropriate organization according to the environment conditions and their unpredictable changes. The main goal of this paper is to propose an alternative to implement and manage autonomous self-organizing systems. This proposal focuses on the well known Scout Movement or Scouting which has been a very successful youth movement in which the self-organization of its members can be observed. The results presented in this paper show that the Scouting principles can be used to automatize a self-adaption mechanism capable to reorganize a MAS along with the discussion of few patents associated with Self- Organizing Multi-Agent Systems.

The Hurst exponent is an index of fundamental importance in the analysis of the long range dependence features of observable time-series. As such, it has been estimated and analyzed in an astonishing number of physical systems. Over the time, various estimation methods as well as generalizations have been suggested and discussed: we therein judge straightforward to review the most important ones. In addition, we offer some insights on recent literature evolution and on patents that address practical implementation of the Hurst exponent.

Proteases play vitally important regulatory roles in life cycle and represent main potential targets for medical intervention. Different types of proteases perform different functions with different biological processes. Therefore, it is highly desired to develop a fast and reliable approach of calculation for identifying the proteases types. From the protein sequence, 20 amino acids are reclassified into 9-gram encoding according to their biophysical and biochemical properties. Using 9 encoding compositions, increment of diversity and low-frequency power spectral density to extract the information of protease sequence, a support vector machine is applied to predict protease types. By the jackknife test, success rates of our algorithm are higher than other methods. In this paper, three recent patents are used.

Reconfigurable Computing systems are the systems that allow computations to be performed in hardware while retaining the flexibility of software. Reconfigurable Computing provides a great potential to accelerate applications such as image processing, sequence searching and matching, encryption, decryption, DNA sequence matching and other compute intensive applications. In this paper we provide an overview of reconfigurable computing architectures, run-time operation, software designs, application environments, and applications and finally the benefits and challenges of it. The paper also portrays some of the patents in the field of reconfigurable computing.

Support vector machine (SVM) which can solve the problem of ‘over-fitting’, local extremum has great application perspective in fault diagnosis of power transformer. However, classification ability of SVM is influenced by the training sample including excessive attribute. In order to solve this difficulty, a new fault diagnosis method for power transformer based on K-mean clustering algorithm (KMC), rough sets(RS) and support vector machine(SVM) is presented in this paper. K-mean clustering algorithm is used to gain discrete data in diagnostic decision table, and decision rule is gained by rough sets. Sample attribute is simplified to construct optimal SVM model through information reduction approach of RS, then the optimal SVM model is used for fault diagnosis of power transformer efficiently and exactly. Finally, the effectiveness and correctness of this method are validated by the result of fault diagnosis examples.