Analyzing temporal patterns of topic diversity using graph clustering

The Journal of Supercomputing - During a disaster, social media can be both a source of help and of danger: Social media has a potential to diffuse rumors, and officials involved in disaster...     

Polynomial summaries of positive semidefinite kernels

Polynomials have proven to be useful tools to tailor generic kernels to specific applications. Nevertheless, we had only restricted knowledge for selecting fertile polynomials which consistently produce positive semidefinite kernels. For example, the well-known polynomial kernel can only take advantage of a very narrow range of polynomials, that is, the univariate polynomials with positive coefficients. This restriction not only hinders intensive exploitation of the flexibility of the kernel method, but also causes misuse of indefinite kernels. Our main theorem significantly relaxes the restriction by asserting that a polynomial consistently produces positive semidefinite kernels, if it has a positive semidefinite coefficient matrix. This sufficient condition is quite natural, and hence, it can be a good characterization of the fertile polynomials. In fact, we prove that the converse of the assertion of the theorem also holds true in the case of degree 1. We also prove the effectiveness of our main theorem by showing three corollaries relating to certain applications known in the literature: the first and second corollaries, respectively, give generalizations of the polynomial kernel and the principal-angle (determinant) kernel. The third corollary shows extended and corrected sufficient conditions for the codon-improved kernel and the weighted-degree kernel with shifts to be positive semidefinite.     

Mechanical Characterization of Amyloid Fibrils Using Coarse‐Grained Normal Mode Analysis

Recent experimental studies have shown that amyloid fibril formed by aggregation of β peptide exhibits excellent mechanical properties comparable to other protein materials such as actin filaments and microtubules. These excellent mechanical properties of amyloid fibrils are related to their functional role in disease expression. This indicates the necessity of understanding how an amyloid fibril achieves the remarkable mechanical properties through self‐aggregation with structural hierarchy. However, the structure‐property–function relationship still remains elusive. In this work, the mechanical properties of human islet amyloid polypeptide (hIAPP) are studied with respect to its structural hierarchies and structural shapes by coarse‐grained normal mode analysis. The simulation shows that hIAPP fibril can achieve the excellent bending rigidity via specific aggregation patterns such as antiparallel stacking of β peptides. Moreover, the length‐dependent mechanical properties of amyloids are found. This length‐dependent property has been elucidated from a Timoshenko beam model that takes into account the shear effect on the bending of amyloids. In summary, the study sheds light on the importance of not only the molecular architecture, which encodes the mechanical properties of the fibril, but also the shear effect on the mechanical (bending) behavior of the fibril.     

Finite size effect on nanomechanical mass detection: the role of surface elasticity

Nanomechanical resonators have recently been highlighted because of their remarkable ability to perform both sensing and detection. Since the nanomechanical resonators are characterized by a large surface-to-volume ratio, it is implied that the surface effect plays a substantial role on not only the resonance but also the sensing performance of nanomechanical resonators. In this work, we have studied the role of surface effect on the detection sensitivity of a nanoresonator that undergoes either harmonic vibration or nonlinear oscillation based on the continuum elastic model such as an elastic beam model. It is shown that the surface effect makes an impact on both harmonic resonance and nonlinear oscillations, and that the sensing performance is dependent on the surface effect. Moreover, we have also investigated the surface effect on the mechanical tuning of resonance and sensing performance. It is interestingly found that the mechanical tuning of resonance is independent of the surface effect, while the mechanical tuning of sensing performance is determined by the surface effect. Our study sheds light on the importance of the surface effect on the sensing performance of nanoresonators.     

Role of Polymeric Metal Nucleation Inducers in Fabricating Large‐Area,Flexible, and Transparent Electrodes for Printable Electronics

The advent of special types of transparent electrodes, known as “ultrathin metal electrodes,” opens a new avenue for flexible and printable electronics based on their excellent optical transparency in the visible range while maintaining their intrinsic high electrical conductivity and mechanical flexibility. In this new electrode architecture, introducing metal nucleation inducers (MNIs) on flexible plastic substrates is a key concept to form high‐quality ultrathin metal films (thickness ≈ 10 nm) with smooth and continuous morphology. Herein, this paper explores the role of “polymeric” MNIs in fabricating ultrathin metal films by employing various polymers with different surface energies and functional groups. Moreover, a scalable approach is demonstrated using the ionic self‐assembly on typical plastic substrates, yielding large‐area electrodes (21 × 29.7 cm²) with high optical transmittance (>95%), low sheet resistance (<10 Ω sq^?1), and extreme mechanical flexibility. The results demonstrate that this new class of flexible and transparent electrodes enables the fabrication of efficient polymer light‐emitting diodes.     

A Fast and Accurate Feature Selection Algorithm Based on Binary Consistency Measure

Consistency‐based feature selection is an important category of feature selection research, and its advantage over other categories is due to consistency measures used to include the effect of interaction among features into evaluation of relevance of features. Even if features individually appear irrelevant to class labels, they can collectively show strong relevance. In such cases, we say that the features interact with each other. Consistency measures, in this regard, evaluate the collective relevance of a set of features and has been intuitively understood as a metric to measure a distance of an arbitrary feature set from the state of being consistent: A set of features is said to be consistent if, and only if, they as a whole determine class labels. In history, the binary consistency measure, which returns the value 1 if the feature set is consistent and 0 otherwise, was the first consistency measure introduced, and many advanced measures followed. The problem of the binary measure consists in the fact that it always returns 1 if a data set includes no consistent feature set. The measures that followed have solved this problem but sacrificed time efficiency of evaluation. Therefore, feature selection leveraging these measures are not fast enough to apply to large data sets. In this article, we aim to improve time efficiency of consistency‐based feature selection. To achieve the goal, we propose a new idea, which we call data set denoising: We eliminate examples which are viewed as noises from a data set until the data set becomes to include consistent feature sets and then apply the binary measure to find an appropriate feature set that is consistent. In our evaluation through intensive experiments, CWC , a new algorithm that implements data set denoising outperformed in both time efficiency and accuracy the benchmark consistency‐based algorithms. Specifically, CWC was about 31 times faster than the LCC that had been known as the fastest in the literature. Furthermore, in a comparison including feature selection algorithms that are not consistency‐based, CWC has turned out to be one of the fastest and the most accurate feature selection algorithms.     

Nonlinear vibration behavior of graphene resonators and their applications in sensitive mass detection

ABSTRACT: Graphene has received significant attention due to its excellent mechanical properties, which has resulted in the emergence of graphene-based nano-electro-mechanical system such as nanoresonators. The nonlinear vibration of a graphene resonator and its application to mass sensing (based on nonlinear oscillation) have been poorly studied, although a graphene resonator is able to easily reach the nonlinear vibration. In this work, we have studied the nonlinear vibration of a graphene resonator driven by a geometric nonlinear effect due to an edge-clamped boundary condition using a continuum elastic model such as a plate model. We have shown that an in-plane tension can play a role in modulating the nonlinearity of a resonance for a graphene. It has been found that the detection sensitivity of a graphene resonator can be improved by using nonlinear vibration induced by an actuation force-driven geometric nonlinear effect. It is also shown that an in-plane tension can control the detection sensitivity of a graphene resonator that operates both harmonic and nonlinear oscillation regimes. Our study suggests the design principles of a graphene resonator as a mass sensor for developing a novel detection scheme using graphene-based nonlinear oscillators.     

A 3-D Coarser-Grained Computational Model for Simulating Large Protein Dynamics

Protein dynamics is essential for gaining insight into biological functions of proteins. Although protein dynamics is well delineated by molecular model, the molecular model is computationally prohibited for simulating large protein structures. In this work, we provide the three-dimensional coarser-grained anisotropic model (CGAM), which is based on model reduction applicable to large protein structures. It is shown that CGAM achieves the fast computation on low-frequency modes, quantitatively comparable to original structural model such as elastic network model (ENM). This indicates that the CGAM by model reduction method enable us to understand the functional motion of large proteins with remarkable computational efficiency.