Current Issue : April - June Volume : 2015 Issue Number : 2 Articles : 4 Articles
Computational modelling of biochemical systems\nbased on top-down and bottom-up approaches has been\nwell studied over the last decade. In this research, after illustrating\nhow to generate atomic components by a set of given\nreactants and two user pre-defined component patterns, we\npropose an integrative top-down and bottom-up modelling\napproach for stepwise qualitative exploration of interactions\namong reactants in biochemical systems. Evolution strategy\nis applied to the top-down modelling approach to compose\nmodels, and simulated annealing is employed in the\nbottom-up modelling approach to explore potential interactions\nbased on models constructed from the top-down modelling\nprocess. Both the top-down and bottom-up approaches\nsupport stepwise modular addition or subtraction for the\nmodel evolution. Experimental results indicate that our modelling\napproach is feasible to learn the relationships among\nbiochemical reactants qualitatively. In addition, hidden reactants\nof the target biochemical system can be obtained by\ngenerating complex reactants in corresponding composed\nmodels. Moreover, qualitatively learned models with inferred reactants and alternative topologies can be used for further\nweb-lab experimental investigations by biologists of interest,\nwhich may result in a better understanding of the system....
In this paper, hybrid ant colony optimization\n(HAntCO) approach in solving multi-skill resource-constrained\nproject scheduling problem (MS-RCPSP) has been presented.\nWe have proposed hybrid approach that links classical\nheuristic priority rules for project scheduling with ant\ncolony optimization (ACO). Furthermore, a novel approach\nfor updating pheromone value has been proposed based on\nboth the best and worst solutions stored by ants. The objective\nof this paper is to research the usability and robustness\nof ACO and its hybrids with priority rules in solving MSRCPSP.\nExperiments have been performed using artificially\ncreated dataset instances based on real-world ones. We published\nthose instances that can be used as a benchmark. Presented\nresults show that ACO-based hybrid method is an\nefficient approach. More directed search process by hybrids\nmakes this approach more stable and provides mostly better\nresults than classical ACO....
Biogeography-based optimization (BBO) is a\nrelatively new heuristic method, where a population of\nhabitats (solutions) are continuously evolved and improved\nmainly by migrating features from high-quality solutions to\nlow-quality ones. In this paper we equip BBO with local\ntopologies, which limit that the migration can only occur\nwithin the neighborhood zone of each habitat. We develop\nthree versions of localized BBO algorithms, which use three\ndifferent local topologies namely the ring topology, the\nsquare topology, and the random topology respectively. Our\napproach is quite easy to implement, but it can effectively\nimprove the search capability and prevent the algorithm from\nbeing trapped in local optima. We demonstrate the effectiveness\nof our approach on a set of well-known benchmark\nproblems.We also introduce the local topologies to a hybrid\nDE/BBO method, resulting in three localized DE/BBO algorithms,\nand show that our approach can improve the performance\nof the state-of-the-art algorithm as well....
One of the most important challenges for\nmachine learning community is to develop efficient classifiers\nwhich are able to cope with data streams, especially\nwith the presence of the so-called concept drift. This phenomenon\nis responsible for the change of classification task\ncharacteristics, and poses a challenge for the learning model\nto adapt itself to the current state of the environment. So there\nis a strong belief that one-class classification is a promising\nresearch direction for data stream analysisââ?¬â?it can be used for\nbinary classification without an access to counterexamples,\ndecomposing a multi-class data stream, outlier detection or\nnovel class recognition. This paper reports a novel modification\nof weighted one-class support vector machine, adapted\nto the non-stationary streaming data analysis. Our proposition\ncan deal with the gradual concept drift, as the introduced\none-class classifier model can adapt its decision boundary to\nnew, incoming data and additionally employs a forgetting\nmechanism which boosts the ability of the classifier to follow\nthe model changes. In this work, we propose several\ndifferent strategies for incremental learning and forgetting,\nand additionally we evaluate them on the basis of several real\ndata streams. Obtained results confirmed the usability of proposed\nclassifier to the problem of data stream classification\nwith the presence of concept drift. Additionally, implemented forgetting mechanism assures the limited memory consumption,\nbecause only quite new and valuable examples should\nbe memorized....
Loading....