Chaos-based learning is a popular method that combines with optimized techniques to achieve the best fitness values by preventing early convergence and improving initial conditions. Chaos enhances the search process, ...
详细信息
Chaos-based learning is a popular method that combines with optimized techniques to achieve the best fitness values by preventing early convergence and improving initial conditions. Chaos enhances the search process, while opposition learning helps improve the population development matrix, ultimately boosting result quality. This paper introduces two new concepts in machine learning. It proposes an improved Opposition and Chaos-based Water Cycle Algorithm (OCWCA) to effectively achieve the best fitness value through cost calculation and to validate the results of objective functions on 15 engineering benchmark problems with constraints. The Water Cycle Algorithm (WCA), a hydrology-based method, provides global search locations based on the flow of streams and rivers toward the sea using predefined control parameters to generate a population matrix. A convergence plot for OCWCA compared to WCA demonstrates significant improvement in results, incorporating learnings from the WCA method and yielding the best fitness data.
We explore an approach to full-body motion editing with linear motion models, prioritized constraint-based optimization and latent-space interpolation. By exploiting the mathematical connections between linear motion ...
详细信息
We explore an approach to full-body motion editing with linear motion models, prioritized constraint-based optimization and latent-space interpolation. By exploiting the mathematical connections between linear motion models and prioritized inverse kinematics (PIK), we formulate and solve the motion editing problem as an optimization function whose differential structure is rich enough to efficiently optimize user-specified constraints within the latent motion space. Performing motion editing within latent motion spaces has the advantage of handling pose transitions and consequently motion flow by construction from single key-frame editing. To handle motion adjustments from multiple key-frame and trajectory constraints, we developed a latent-space interpolation technique by exploiting spline functions. Such an approach handles per-frame adjustments generating smooth animations, while avoiding the computational expense of joint space interpolations. We demonstrate the usefulness of this approach by editing and generating full-body reaching and walking jump animations in challenging environment scenarios.
In learning by demonstration, the generalization of motion trajectories far away from the set of demonstrations is often limited by the dependency of the learned models on arbitrary coordinate references. Trajectory s...
详细信息
In learning by demonstration, the generalization of motion trajectories far away from the set of demonstrations is often limited by the dependency of the learned models on arbitrary coordinate references. Trajectory shape descriptors have the potential to remove these dependencies by representing demonstrated trajectories in a coordinate-free way. This paper proposes a constraint-based optimization framework to generalize demonstrated rigid-body motion trajectories to new situations starting from the shape descriptor of the demonstration. Experimental results indicate excellent generalization capabilities showing how, starting from only a single demonstration, new trajectories are easily generalized to novel situations anywhere in task space, such as new initial or target positions and orientations, while preserving similarity with the demonstration. The results encourage the use of trajectory shape descriptors in learning by demonstration to reduce the number of required demonstrations. (C) 2019 Elsevier B.V. All rights reserved.
Streptomyces coelicolor is a model organism for the Actinobacteria, a phylum known to produce an extensive range of different bioactive compounds that include antibiotics currently used in the clinic. Biosynthetic gen...
详细信息
Streptomyces coelicolor is a model organism for the Actinobacteria, a phylum known to produce an extensive range of different bioactive compounds that include antibiotics currently used in the clinic. Biosynthetic gene clusters discovered in genomes of other Actinobacteria can be transferred to and expressed in S. coelicolor, making it a factory for heterologous production of secondary metabolites. Genome-scale metabolic reconstructions have successfully been used in several biotechnology applications to facilitate the over-production of target metabolites. Here, the authors present iKS1317, the most comprehensive and accurate reconstructed genome-scale metabolic model (GEM) for S. coelicolor. The model reconstruction is based on previous models, publicly available databases, and published literature and includes 1317 genes, 2119 reactions, and 1581 metabolites. It correctly predicts wild-type growth in 96.5% of the evaluated growth environments and gene knockout predictions in 78.4% when comparing with observed mutant growth phenotypes, with a total accuracy of 83.3%. However, using a minimal nutrient environment for the gene knockout predictions, iKS1317 has an accuracy of 87.1% in predicting mutant growth phenotypes. Furthermore, we used iKS1317 and existing strain design algorithms to suggest robust gene-knockout strategies to increase the production of acetyl-CoA. Since acetyl-CoA is the most important precursor for polyketide antibiotics, the suggested strategies may be implemented in vivo to improve the function of S. coelicolor as a heterologous expression host.
In the last decade, plant genome-scale modeling has developed rapidly and modeling efforts have advanced from representing metabolic behavior of plant heterotrophic cell suspensions to studying the complex interplay o...
详细信息
In the last decade, plant genome-scale modeling has developed rapidly and modeling efforts have advanced from representing metabolic behavior of plant heterotrophic cell suspensions to studying the complex interplay of cell types, tissues, and organs. A crucial driving force for such developments is the availability and integration of “omics” data (e.g., transcriptomics, proteomics, and metabolomics) which enable the reconstruction, extraction, and application of context-specific metabolic networks. In this chapter, we demonstrate a workflow to integrate gas chromatography coupled to mass spectrometry (GC-MS)-based metabolomics data of tomato fruit pericarp (flesh) tissue, at five developmental stages, with a genome-scale reconstruction of tomato metabolism. This method allows for the extraction of context-specific networks reflecting changing activities of metabolic pathways throughout fruit development and maturation. less
暂无评论