The silk moth （Bombyx mori） exhibits efficient Chemical Plume Tracing （CPT）, which is ideal for biomimetics. However, there is insufficient quantitative understanding of its CPT behavior. We propose a hierarchical...The silk moth （Bombyx mori） exhibits efficient Chemical Plume Tracing （CPT）, which is ideal for biomimetics. However, there is insufficient quantitative understanding of its CPT behavior. We propose a hierarchical classification method to segment its natural CPT locomotion and to build its inverse model for detecting stimulus input. This provides the basis for quantitative analysis. The Gaussian mixture model with expectation-maximization algorithm is used first for unsupervised classification to decompose CPT locomotion data into Gaussian density components that represent a set of quantified elemental motions. A heuristic behavioral rule is used to categorize these components to eliminate components that are descriptive of the same motion. Then, the echo state network is used for supervised classification to evaluate segmented elemental motions and to compare CPT locomotion among different moths. In this case, categorized elemental motions are used as the training data to estimate stimulus time. We successfully built the inverse CPT behavioral model of the silk moth to detect stimulus input with good accuracy. The quantitative analysis indicates that silk moths exhibit behavioral singularity and time dependency in their CPT locomotion, which is dominated by its singularity.展开更多
Several decades ago,Profs.Sean Meyn and Lei Guo were postdoctoral fellows at ANU,where they shared interest in recursive algorithms.It seems fitting to celebrate Lei Guo’s 60 th birthday with a review of the ODE Meth...Several decades ago,Profs.Sean Meyn and Lei Guo were postdoctoral fellows at ANU,where they shared interest in recursive algorithms.It seems fitting to celebrate Lei Guo’s 60 th birthday with a review of the ODE Method and its recent evolution,with focus on the following themes:The method has been regarded as a technique for algorithm analysis.It is argued that this viewpoint is backwards:The original stochastic approximation method was surely motivated by an ODE,and tools for analysis came much later(based on establishing robustness of Euler approximations).The paper presents a brief survey of recent research in machine learning that shows the power of algorithm design in continuous time,following by careful approximation to obtain a practical recursive algorithm.While these methods are usually presented in a stochastic setting,this is not a prerequisite.In fact,recent theory shows that rates of convergence can be dramatically accelerated by applying techniques inspired by quasi Monte-Carlo.Subject to conditions,the optimal rate of convergence can be obtained by applying the averaging technique of Polyak and Ruppert.The conditions are not universal,but theory suggests alternatives to achieve acceleration.The theory is illustrated with applications to gradient-free optimization,and policy gradient algorithms for reinforcement learning.展开更多
文摘The silk moth （Bombyx mori） exhibits efficient Chemical Plume Tracing （CPT）, which is ideal for biomimetics. However, there is insufficient quantitative understanding of its CPT behavior. We propose a hierarchical classification method to segment its natural CPT locomotion and to build its inverse model for detecting stimulus input. This provides the basis for quantitative analysis. The Gaussian mixture model with expectation-maximization algorithm is used first for unsupervised classification to decompose CPT locomotion data into Gaussian density components that represent a set of quantified elemental motions. A heuristic behavioral rule is used to categorize these components to eliminate components that are descriptive of the same motion. Then, the echo state network is used for supervised classification to evaluate segmented elemental motions and to compare CPT locomotion among different moths. In this case, categorized elemental motions are used as the training data to estimate stimulus time. We successfully built the inverse CPT behavioral model of the silk moth to detect stimulus input with good accuracy. The quantitative analysis indicates that silk moths exhibit behavioral singularity and time dependency in their CPT locomotion, which is dominated by its singularity.
基金ARO W911NF1810334NSF under EPCN 1935389the National Renewable Energy Laboratory(NREL)。
文摘Several decades ago,Profs.Sean Meyn and Lei Guo were postdoctoral fellows at ANU,where they shared interest in recursive algorithms.It seems fitting to celebrate Lei Guo’s 60 th birthday with a review of the ODE Method and its recent evolution,with focus on the following themes:The method has been regarded as a technique for algorithm analysis.It is argued that this viewpoint is backwards:The original stochastic approximation method was surely motivated by an ODE,and tools for analysis came much later(based on establishing robustness of Euler approximations).The paper presents a brief survey of recent research in machine learning that shows the power of algorithm design in continuous time,following by careful approximation to obtain a practical recursive algorithm.While these methods are usually presented in a stochastic setting,this is not a prerequisite.In fact,recent theory shows that rates of convergence can be dramatically accelerated by applying techniques inspired by quasi Monte-Carlo.Subject to conditions,the optimal rate of convergence can be obtained by applying the averaging technique of Polyak and Ruppert.The conditions are not universal,but theory suggests alternatives to achieve acceleration.The theory is illustrated with applications to gradient-free optimization,and policy gradient algorithms for reinforcement learning.
