Logistic label propagation

In this paper, we propose a novel method for semi-supervised learning, called logistic label propagation (LLP). The proposed method employs the logistic function to classify input pattern vectors, similarly to logistic regression. To cope with unlabeled samples as well as labeled ones in the semi-supervised learning framework, the logistic functions are learnt by using similarities between samples in a manner similar to label propagation. In the proposed method, these two methods of logistic regression and label propagation are effectively incorporated in terms of posterior probabilities. LLP estimates the labels of input samples by using the learnt logistic function, whereas the method of label propagation has to optimize the whole labels whenever an input sample comes. In addition, we suggest the way to provide proper parameter setting and initialization, which frees the users from determining a parameter value in trial and error. In experiments on classification (estimating labels) in the semi-supervised learning framework, the proposed method exhibits favorable performances compared to the other methods.     

Experimental investigation of distributed navigation and collision avoidance for a robotic swarm

Artificial Life and Robotics - In this study, we focus on the navigation of a robotic swarm and perform an experimental investigation. A decentralized control method using only local information...     

Fluorophore-Decorated Mie Resonant Silicon Nanosphere for Scattering/Fluorescence Dual-Mode Imaging

Inorganic nanoparticles with multiple functions have been attracting attention as multimodal nanoprobes in bioimaging, biomolecule detection, and medical diagnosis and treatment. A drawback of conventional metallic nanoparticle-based nanoprobes is the Ohmic losses that lead to fluorescence quenching of attached molecules and local heating under light irradiation. Here, metal-free nanoprobes capable of scattering/fluorescence dual-mode imaging are developed. The nanoprobes are composed of a silicon nanosphere core having efficient Mie scattering in the visible to near infrared range and a fluorophore doped silica shell. The dark-field scattering and photoluminescence images/spectra for nanoprobes made from different size silicon nanospheres and different kinds of fluorophores are studied by single particle spectroscopy. The fluorescence spectra are strongly modified by the Mie modes of a silicon nanosphere core. By comparing scattering and fluorescence spectra and calculated Purcell factors, the fluorescence enhancement factor is quantitatively discussed. In vitro scattering/fluorescence imaging studies on human cancer cells demonstrate that the developed nanoparticles work as scattering/fluorescence dual-mode imaging nanoprobes.     

Melt-quenched oxide glasses with ultrahigh Young's modulus and small thermal expansion coefficient

The well-known Makishima–Mackenzie relationship, consisting of two terms of the dense packing structure and dissociation energy regarding bonding in constituent oxides, enables fabricating oxide glasses with ultrahigh Young's modulus (∼140 GPa) and a small coefficient of thermal expansion (CTE) (∼4 ppm/K). The effects of increasing MgO and Ta₂O₅ contents in an MgO–Ta₂O₅–Al₂O₃–SiO₂–B₂O₃ glass system using a conventional melt-quenching method are revealed. The essential oxides of Al₂O₃ and Ta₂O₅ are primarily suitable for dense packing structures dominated by a large coordination number of oxygens. The substitution of CaO by MgO results in high dissociation energy when the glass composition falls in the peraluminous regime (Al₂O₃/[MgO + CaO] > 1). A small CTE is realized by increasing the molar ratio of Al₂O₃/MgO. According to magic-angle spinning-nuclear magnetic resonance spectra, mechanically and thermally functional oxide glasses depend on their structures. These findings facilitate the development of glass substrate applications without thermal dilatation.