New Interfaces and Approaches to Machine Learning When Classifying Gestures within Music

Interactive music uses wearable sensors (i.e., gestural interfaces—GIs) and biometric datasets to reinvent traditional human–computer interaction and enhance music composition. In recent years, machine learning (ML) has been important for the artform. This is because ML helps process complex biometric datasets from GIs when predicting musical actions (termed performance gestures). ML allows musicians to create novel interactions with digital media. Wekinator is a popular ML software amongst artists, allowing users to train models through demonstration. It is built on the Waikato Environment for Knowledge Analysis (WEKA) framework, which is used to build supervised predictive models. Previous research has used biometric data from GIs to train specific ML models. However, previous research does not inform optimum ML model choice, within music, or compare model performance. Wekinator offers several ML models. Thus, we used Wekinator and the Myo armband GI and study three performance gestures for piano practice to solve this problem. Using these, we trained all models in Wekinator and investigated their accuracy, how gesture representation affects model accuracy and if optimisation can arise. Results show that neural networks are the strongest continuous classifiers, mapping behaviour differs amongst continuous models, optimisation can occur and gesture representation disparately affects model mapping behaviour; impacting music practice.     

On the use of the resting potential and level set methods for identifying ischemic heart disease: An inverse problem

The electrical activity in the heart is modeled by a complex, nonlinear, fully coupled system of differential equations. Several scientists have studied how this model, referred to as the bidomain model, can be modified to incorporate the effect of heart infarctions on simulated ECG (electrocardiogram) recordings.We are concerned with the associated inverse problem; how can we use ECG recordings and mathematical models to identify the position, size and shape of heart infarctions? Due to the extreme CPU efforts needed to solve the bidomain equations, this model, in its full complexity, is not well-suited for this kind of problems. In this paper we show how biological knowledge about the resting potential in the heart and level set techniques can be combined to derive a suitable stationary model, expressed in terms of an elliptic PDE, for such applications. This approach leads to a nonlinear ill-posed minimization problem, which we propose to regularize and solve with a simple iterative scheme.Finally, our theoretical findings are illuminated through a series of computer simulations for an experimental setup involving a realistic heart in torso geometry. More specifically, experiments with synthetic ECG recordings, produced by solving the bidomain model, indicate that our method manages to identify the physical characteristics of the ischemic region(s) in the heart. Furthermore, the ill-posed nature of this inverse problem is explored, i.e. several quantitative issues of our scheme are explored.     

Tomography Methods for Vector Field Study Using Space-Distributed Fiber Optic Sensors with Integral Sensitivity

The problem of tomographic reconstruction of vector physical fields is studied. This problem can be solved by using fiber optic measuring lines (MLs) of special shape. In the case that the ML output signal is proportional to the vector's projection, the ML must be shaped like a narrow loop. This problem can be solved by means of the integral theorem. If an ML output signal is proportional to projection of a vector derivative with respect to ML direction, the ML with a step shape can be used. In this case the potential component of a vector field can be reconstructed. This approach can be applied to research on distributions of electromagnetic, deforming, and other vector fields and can be used for developing systems to monitor vector physical fields.     

Latency optimization for Federated Learning over Wireless Power Transfer

Federal Learning (FL) is an emerging technology in the field of machine learning (ML). Compared with traditional ML, FL is an attractive method to deal with data security issues of the user-side. So that FL can realizes its full potential in terms of low latency and high energy efficiency (EE), this paper introduces a new framework: In the wireless communication network scenario, we propose an FL architecture based on Wireless Power Transfer (WPT). By combining WPT technology and FL, we can realize green wireless communication under the premise of ensuring user privacy. We formulate a joint calculation and communication optimization problem to optimize the latency of local calculation, uplink and downlink transmission without consuming user-side energy. The problem formulas listed according to the optimization problem are non-convex. They are first transformed into convex form, and then a low-complexity iterative algorithm is used to solve them optimally. Simulations show that our proposed FL method design has achieved a significant performance improvement over other benchmark tests.     

Surface structures of atomic hydrogen adsorbed on Cu(1 1 1) surface studied by density-functional-theory calculations

The surface structures of atomic hydrogen adsorbed on Cu(1 1 1) surface have been studied theoretically by using density-functional-theory calculations. The results show that 0.67 ML hydrogen adsorbed on threefold hollow sites forming (3 × 1) superstructure and 0.5 ML hydrogen adsorbed on threefold hollow sites forming (2 × 2)-2H superstructure with central H at trigonal sites induce most significant substrate reconstructions and that fits best the observed (3 × 3) and (2 × 2) LEED patterns, respectively. The potential energies for the hydrogen in these two models are also lower than those in other competing models. Accordingly, these two models are the most preferable structures for 0.5-0.67 ML and 0.3-0.5 ML hydrogen adsorbed on the Cu(1 1 1) surface. In addition, the calculations also suggest that the lateral H-H interaction is not of simple repulsion and how the adsorbed hydrogen is arrayed is important in modifying the adsorption energy.     

Why error-prone quantum measurements have outcomes

To solve the quantum measurement problem it is necessary to construct quantum mechanical models of measurement interactions to show why properly conducted measurements always yield definite outcomes. The main barrier to a solution has been the interpretive principle that a quantum system has a definite value for an observable only if it may be described by a quantum eigenstate of the corresponding operator. I have recently proposed a solution to the measurement problem based on alternative interpretive principles. The present paper defends this proposal against recent criticisms which seek to show that it fails to solve the problem unless quantum measurements meet highly idealized conditions which no actual measurement could hope to meet. Several models of error-prone measurements are shown to lead to definite outcomes, and a general defense of the appropriateness of these models is sketched.     

Security concerns on machine learning solutions for 6G networks in mmWave beam prediction

6G – sixth generation – is the latest cellular technology currently under development for wireless communication systems. In recent years, machine learning (ML) algorithms have been applied widely in various fields, such as healthcare, transportation, energy, autonomous cars, and many more. Those algorithms have also been used in communication technologies to improve the system performance in terms of frequency spectrum usage, latency, and security. With the rapid developments of ML techniques, especially deep learning (DL), it is critical to consider the security concern when applying the algorithms. While ML algorithms offer significant advantages for 6G networks, security concerns on artificial intelligence (AI) models are typically ignored by the scientific community so far. However, security is also a vital part of AI algorithms because attackers can poison the AI model itself. This paper proposes a mitigation method for adversarial attacks against proposed 6G ML models for the millimeter-wave (mmWave) beam prediction using adversarial training. The main idea behind generating adversarial attacks against ML models is to produce faulty results by manipulating trained DL models for 6G applications for mmWave beam prediction. We also present a proposed adversarial learning mitigation method’s performance for 6G security in mmWave beam prediction application a fast gradient sign method attack. The results show that the defended model under attack’s mean square errors (i.e., the prediction accuracy) are very close to the undefended model without attack.     

Partitioning networks into clusters and residuals with average association

We investigate the problem of detecting clusters exhibiting higher-than-average internal connectivity in networks of interacting systems. We show how the average association objective formulated in the context of spectral graph clustering leads naturally to a clustering strategy where each system is assigned to at most one cluster. A residual set is formed of the systems that are not members of any cluster. Maximization of the average association objective leads to a discrete optimization problem, which is difficult to solve, but a relaxed version can be solved using an eigendecomposition of the connectivity matrix. A simple approach to extracting clusters from a relaxed solution is described and developed by applying a variance maximizing solution to the relaxed solution, which leads to a method with increased accuracy and sensitivity. Numerical studies of theoretical connectivity models and of synchronization clusters in a lattice of coupled Lorenz oscillators are conducted to show the efficiency of the proposed approach. The method is applied to an experimentally obtained human resting state functional magnetic resonance imaging dataset and the results are discussed.     

Machine learning identification of symmetrized base states of Rydberg atoms

Studying the complex quantum dynamics of interacting many-body systems is one of the most challenging areas in modern physics. Here, we use machine learning (ML) models to identify the symmetrized base states of interacting Rydberg atoms of various atom numbers (up to six) and geometric configurations. To obtain the data set for training the ML classifiers, we generate Rydberg excitation probability profiles that simulate experimental data by utilizing Lindblad equations that incorporate laser intensities and phase noise. Then, we classify the data sets using support vector machines (SVMs) and random forest classifiers (RFCs). With these ML models, we achieve high accuracy of up to 100% for data sets containing only a few hundred samples, especially for the closed atom configurations such as the pentagonal (five atoms) and hexagonal (six atoms) systems. The results demonstrate that computationally cost-effective ML models can be used in the identification of Rydberg atom configurations.     

First-order system least squares and the energetic variational approach for two-phase flow

This paper develops a first-order system least-squares (FOSLS) formulation for equations of two-phase flow. The main goal is to show that this discretization, along with numerical techniques such as nested iteration, algebraic multigrid, and adaptive local refinement, can be used to solve these types of complex fluid flow problems. In addition, from an energetic variational approach, it can be shown that an important quantity to preserve in a given simulation is the energy law. We discuss the energy law and inherent structure for two-phase flow using the Allen–Cahn interface model and indicate how it is related to other complex fluid models, such as magnetohydrodynamics. Finally, we show that, using the FOSLS framework, one can still satisfy the appropriate energy law globally while using well-known numerical techniques.     

LWDG Method for a Multi-Class Traffic Flow Model on an Inhomogeneous Highway

In this paper, we apply the discontinuous Galerkin method with Lax-Wendroff type time discretizations (LWDG) using the weighted essentially non-oscillatory (WENO) limiter to solve a multi-class traffic flow model for an inhomogeneous highway, which is a kind of hyperbolic conservation law with spatially varying fluxes. The numerical scheme is based on a modified equivalent system which is written as a "standard" hyperbolic conservation form. Numerical experiments for both the Riemann problem and the traffic signal control problem are presented to show the effectiveness of these methods.     

Adjusting the properties of the photon generated via an optical parametric oscillator by using a pulse pumped laser

The cavity-enhanced spontaneous parametric down-conversion far below threshold can be used to generate a narrow-band photon pair efficiently. Previous experiments on the cavity-enhanced spontaneous parametric down-conversion almost always utilize continuous wave pump light, but the pulse pumped case is rarely reported. One disadvantage of the continuous wave case is that the photon pair is produced randomly within the coherence time of the pump, which limits its application in the quantum information realm. However, a pulse pump can help to solve this problem. In this paper, we theoretically analyze pulse pumped cavity-enhanced spontaneous parametric down-conversion in detail and show how the pump pulse affects the multi-photon interference visibility, two-photon waveform, joint spectrum and spectral brightness.     

不确定海洋环境中的模态子空间重构稳健定位方法

实际的海洋是一个不确定的声传播环境,常规的匹配场方法在进行目标定位时会遇到环境失配的问题,导致定位性能下降.在不确定的海洋环境中,声场传播中的一部分简正波模态受到声场不确定性的影响较小.基于此,本文提出了一种模态子空间重构的稳健定位方法.该方法使用稳定的模态来重构拷贝场向量,相比于常规匹配场定位方法中使用全阶模态来构造拷贝场向量,其定位结果更加稳健.利用计算机仿真数据和海试数据进行了定位性能分析,并给出了常规匹配场定位方法和稳健最大似然定位方法作为对比.研究结果表明:1)不确定海洋环境中,常规匹配场定位方法即使在较高的信噪比条件下其定位性能也较差.2)模态子空间重构定位方法的性能优于常规匹配场定位方法和稳健最大似然方法.     

蒙特卡罗方法在解微分方程边值问题中的应用

介绍了蒙特卡罗方法的基本原理以及随机数的产生方法。基于蒙特卡罗方法的思想,结合有限差分方法,建立了求解微分方程边值问题的随机概率模型,并以第一类边界条件的拉普拉斯方程和一个给定初值及边界条件的非稳态热传导方程为数值算例,研究了蒙特卡罗方法在求解微分方程边值问题中的应用。结果表明:利用蒙特卡罗方法,不仅可以有效解决给定边界条件的微分方程,对于给定初值条件的微分方程,也可以从时域有限差分方程出发,采用蒙特卡罗方法进行求解。数值模拟和对误差的理论分析均表明,增加蒙特卡罗试验中的模拟粒子点数,可以提高计算结果的精度。     

Machine Learning and Micromagnetic Studies of Magnetization Switching

Magnetization switching is one of the most fundamental topics in the field of magnetism.Machine learning(ML)models of random forest(RF),support vector machine(SVM),deep neural network(DNN)methods are built and trained to classify the magnetization reversal and non-reversal cases of single-domain particle,and the classification performances are evaluated by comparison with micromagnetic simulations.The results show that the ML models have achieved great accuracy and the DNN model reaches the best area under curve(AUC)of 0.997,even with a small training dataset,and RF and SVM models have lower AUCs of 0.964 and 0.836,respectively.This work validates the potential of ML applications in studies of magnetization switching and provides the benchmark for further ML studies in magnetization switching.     

Spatiotemporal growing wave fronts in spatially stable boundary layers

In fluid dynamical systems, it is not known a priori whether disturbances grow either in space or in time or as spatiotemporal structures. For a zero pressure gradient boundary layer (also known as the Blasius boundary layer), it is customary to treat it as a spatial problem, and some limited comparison between prediction and laboratory experiments exist. In the present work, the two-dimensional receptivity problem of a Blasius boundary layer excited by a localized harmonic source is investigated under the general spatiotemporal framework, by using the Bromwich contour integral method. While this approach is seen to be equivalent to the spatial study for unstable systems, here we show for the first time how spatially stable systems show spatiotemporally growing wave fronts.     

A novel location estimation based on pattern matching algorithm in underwater environments

In this paper, we present a novel approach based on pattern recognition to treat the underwater localization. The goal is to achieve underwater localization by the pattern matching algorithm. It should be noted that the reflected signals in underwater environments do not affect our location estimation. Therefore, the underwater localization in this study is straightforward and efficient by using the pattern matching algorithm. We exploit the maximum likelihood (ML) to perform our study. Initially, the underwater signals are collected by the sound receiver at some sampling locations. These signals are suitably processed by the ML models and are stored in database. The test location in real-time is estimated through the database. Experimental results show that good accuracy of positioning can be obtained by proposed schemes. The proposed localization schemes can be applied to many other applications in underwater environments.     

A Schur complement formulation for solving free-boundary,Stefan problems of phase change

A Schur complement formulation that utilizes a linear iterative solver is derived to solve a free-boundary, Stefan problem describing steady-state phase change via the Isotherm–Newton approach, which employs Newton’s method to simultaneously and efficiently solve for both interface and field equations. This formulation is tested alongside more traditional solution strategies that employ direct or iterative linear solvers on the entire Jacobian matrix for a two-dimensional sample problem that discretizes the field equations using a Galerkin finite-element method and employs a deforming-grid approach to represent the melt–solid interface. All methods demonstrate quadratic convergence for sufficiently accurate Newton solves, but the two approaches utilizing linear iterative solvers show better scaling of computational effort with problem size. Of these two approaches, the Schur formulation proves to be more robust, converging with significantly smaller Krylov subspaces than those required to solve the global system of equations. Further improvement of performance are made through approximations and preconditioning of the Schur complement problem. Hence, the new Schur formulation shows promise as an affordable, robust, and scalable method to solve free-boundary, Stefan problems. Such models are employed to study a wide array of applications, including casting, welding, glass forming, planetary mantle and glacier dynamics, thermal energy storage, food processing, cryosurgery, metallurgical solidification, and crystal growth.     

社交网络中热点话题深度挖掘方法及实验测试

互联网大数据时代,社交网络数据交互具有实时性、社会性、复杂性。多元架构数据信息中会存在热点话题数据。传统社交网络热点话题挖掘方法存在热点词条检索速度慢、话题词条层浅、断层数据无法挖掘等一系列问题,如何将社交网络中的热点话题数据进行挖掘,针对这一问题提出大数据信息词条特征比对提取方法,对社交网络中的大数据数据信息词条进行特征显化处理,采用饼图对比方式对特征化词条数据进行网络互交频率展现,采用多维数据获取法,解决社交网络热点话题挖掘中出现的数据阻滞现象,满足社交网络中热点话题深度挖掘的要求。通过仿真实验对提出方法进行效率、准确度、速度测试,实验结果表明,提出方法对社交网络中的热点话题挖掘快捷、高效、实用性强。     

脉冲堆积方式产生整形脉冲的逆问题

在利用光纤脉冲堆积器进行脉冲时间整形时,存在如何确定光纤脉冲堆积器各路的延时、衰减等参数以产生所需脉冲波形的问题.为此分非相干脉冲堆积和部分相干脉冲堆积两种情况,定量分析了堆积脉冲的特性.在此基础上,提出了一种逆算方法,使得在任意给定时间波形情况下,均可以通过该逆算方法,简单而准确地计算出光纤脉冲堆积器参数.研究结果表明,非相干堆积而成的脉冲波形较为平滑,而部分相干堆积而成的脉冲波形则存在一定的调制;按逆算结果堆积而成的脉冲波形与给定波形符合很好.