
The scattering of electromagnetic pulses is described using a nonsingular
boundary integral method to solve directly for the field components in the
frequency domain, and Fourier transform is then used to obtain the complete
spacetime behavior. This approach is stable for wavelengths both small and
large relative to characteristic length scales. Amplitudes and phases of field
values can be obtained accurately on or near material boundaries. Local field
enhancement effects due to multiple scattering of interest to applications in
microphotonics are demonstrated.

We present a boundary integral formulation of electromagnetic scattering by
homogeneous bodies that are characterized by linear constitutive equations in
the frequency domain. By working with the Cartesian components of the electric,
E and magnetic, H fields and with the scalar functions (r*E) and (r*H), the
problem is cast as solving a set of scalar Helmholtz equations for the field
components that are coupled by the usual electromagnetic boundary conditions at
material boundaries. This facilitates a direct solution for E and H rather than
working with surface currents as intermediate quantities in existing methods.
Consequently, our formulation is free of the wellknown numerical instability
that occurs in the zero frequency or long wavelength limit in traditional
surface integral solutions of Maxwell's equations and our numerical results
converge uniformly to the static results in the long wavelength limit.
Furthermore, we use a formulation of the scalar Helmholtz equation that is
expressed as classically convergent integrals and does not require the
evaluation of principal value integrals or any knowledge of the solid angle.
Therefore, standard quadrature and higher order surface elements can readily be
used to improve numerical precision. In addition, near and far field values can
be calculated with equal precision and multiscale problems in which the
scatterers possess characteristic length scales that are both large and small
relative to the wavelength can be easily accommodated. From this we obtain
results for the scattering and transmission of electromagnetic waves at
dielectric boundaries that are valid for any ratio of the local surface
curvature to the wave number. This is a generalization of the familiar Fresnel
formula and Snell's law, valid at planar dielectric boundaries, for the
scattering and transmission of electromagnetic waves at surfaces of arbitrary
curvature.

A boundary integral formulation of electromagnetics that involves only the
components of $\boldsymbol{E}$ and $\boldsymbol{H}$ is derived without the use
of surface currents that appear in the classical PMCHWT formulation. The
kernels of the boundary integral equations for $\boldsymbol{E}$ and
$\boldsymbol{H}$ are nonsingular so that all field quantities at the surface
can be determined to high precision and also geometries with closely spaced
surfaces present no numerical difficulties. Quadratic elements can readily be
used to represent the surfaces so that the surface integrals can be calculated
to higher numerical precision than using planar elements for the same numbers
of degrees of freedom.

Big data can easily be contaminated by outliers or contain variables with
heavytailed distributions, which makes many conventional methods inadequate.
To address this challenge, we propose the adaptive Huber regression for robust
estimation and inference. The key observation is that the robustification
parameter should adapt to the sample size, dimension and moments for optimal
tradeoff between bias and robustness. Our theoretical framework deals with
heavytailed distributions with bounded $(1+\delta)$th moment for any $\delta
> 0$. We establish a sharp phase transition for robust estimation of regression
parameters in both low and high dimensions: when $\delta \geq 1$, the estimator
admits a subGaussiantype deviation bound without subGaussian assumptions on
the data, while only a slower rate is available in the regime $0<\delta< 1$.
Furthermore, this transition is smooth and optimal. In addition, we extend the
methodology to allow both heavytailed predictors and observation noise.
Simulation studies lend further support to the theory. In a genetic study of
cancer cell lines that exhibit heavytailedness, the proposed methods are shown
to be more robust and predictive.

We propose a class of dimension reduction methods for right censored survival
data using a counting process representation of the failure process.
Semiparametric estimating equations are constructed to estimate the dimension
reduction subspace for the failure time model. The proposed method addresses
two fundamental limitations of existing approaches. First, using the counting
process formulation, it does not require any estimation of the censoring
distribution to compensate the bias in estimating the dimension reduction
subspace. Second, the nonparametric part in the estimating equations is
adaptive to the structural dimension, hence the approach circumvents the curse
of dimensionality. Asymptotic normality is established for the obtained
estimators. We further propose a computationally efficient approach that
simplifies the estimation equation formulations and requires only a singular
value decomposition to estimate the dimension reduction subspace. Numerical
studies suggest that our new approaches exhibit significantly improved
performance for estimating the true dimension reduction subspace. We further
conduct a real data analysis on a skin cutaneous melanoma dataset from The
Cancer Genome Atlas. The proposed method is implemented in the R package
"orthoDr".

In our recent works, based on the structural studies on water and interfacial
water (topmost water layer at the solute/water interface), hydration free
energy is derived and utilized to investigate the physical origin of
hydrophobic interactions. In this study, it is extended to investigate the
structural origin of hydration repulsive force. As a solute is embedded into
water, it mainly affects the structure of interfacial water, which is dependent
on the geometric shape of solute. Therefore, hydrophobic interactions may be
related to the surface roughness of solute. According to this study, hydration
repulsive force can reasonably be ascribed to the effects of surface roughness
of solutes on hydrophobic interactions. Additionally, hydration repulsive force
can only be expected as the size of surface roughness being less than Rc
(critical radius), which is in correspondence with the initial solvation
process as discussed in our recent work. Additionally, this can be demonstrated
by potential of mean force (PMF) calculated using molecular dynamics
simulations.

Based on our recent study on physical origin of hydrophobic effects, this is
applied to investigate the dependence of hydrophobic interactions on the solute
size. As two same hydrophobic solutes are dissolved into water, the hydration
free energy is determined, and the critical radius (Rc) is calculated to be 3.2
Angstrom. With reference to the Rc, the dissolved behaviors can be divided into
initial and hydrophobic solvation processes. These can be demonstrated by the
molecular dynamics simulations on C60C60 fullerenes in water, and CH4CH4
molecules in water. In the association of C60 fullerenes in water, with
decreasing the separation between C60 fullerenes, hydrophobic interactions can
be divided into H1w and H2s hydrophobic processes, respectively. In addition,
it can be derived that maximizing hydrogen bonding provides the driving force
in the association of hydrophobic solutes in water.

We establish the counterpart of Hoeffding's lemma for Markov dependent random
variables. Specifically, if a stationary Markov chain $\{X_i\}_{i \ge 1}$ with
invariant measure $\pi$ admits an $\mathcal{L}_2(\pi)$spectral gap
$1\lambda$, then for any bounded functions $f_i: x \mapsto [a_i,b_i]$, the sum
of $f_i(X_i)$ is subGaussian with variance proxy $\frac{1+\lambda}{1\lambda}
\cdot \sum_i \frac{(b_ia_i)^2}{4}$. The counterpart of Hoeffding's inequality
immediately follows. Our results assume none of reversibility, countable state
space and timehomogeneity of Markov chains. They are optimal in terms of the
multiplicative coefficient $(1+\lambda)/(1\lambda)$, and cover Hoeffding's
lemma and inequality for independent random variables as special cases with
$\lambda = 0$. We illustrate the utility of these results by applying them to
six problems in statistics and machine learning. They are linear regression,
lasso regression, sparse covariance matrix estimation with Markovdependent
samples; Markov chain Monte Carlo estimation; respondence driven sampling; and
multiarmed bandit problems with Markovian rewards.

Big data is transforming our world, revolutionizing operations and analytics
everywhere, from financial engineering to biomedical sciences. The complexity
of big data often makes dimension reduction techniques necessary before
conducting statistical inference. Principal component analysis, commonly
referred to as PCA, has become an essential tool for multivariate data analysis
and unsupervised dimension reduction, the goal of which is to find a lower
dimensional subspace that captures most of the variation in the dataset. This
article provides an overview of methodological and theoretical developments of
PCA over the last decade, with focus on its applications to big data analytics.
We first review the mathematical formulation of PCA and its theoretical
development from the view point of perturbation analysis. We then briefly
discuss the relationship between PCA and factor analysis as well as its
applications to large covariance estimation and multiple testing. PCA also
finds important applications in many modern machine learning problems, and we
focus on community detection, ranking, mixture model and manifold learning in
this paper.

In this paper we give a proof of Enomoto's conjecture for graphs of
sufficiently large order. Enomoto's conjecture states that, if $G$ is a graph
of order $n$ with minimum degree $\delta(G)\geq \frac{n}{2}+1$, then for any
pair of vertices $x$, $y$ in $G$, there is a Hamiltonian cycle $C$ of $G$ such
that $d_C(x,y)=\lfloor \frac{n}{2}\rfloor$. The main tools of our proof are
Regularity Lemma of Szemer\'edi and Blowup Lemma of Koml\'os et al.

In this paper, the performance of uplink spectral efficiency in massive
multiple input multiple output (MIMO) over spacially correlated Ricean fading
channel is presented. The maximum ratio combining (MRC) receiver is employed at
the base station (BS) for two different methods of channel estimation. The
first method is based on pilotassisted least minimum mean square error (LMMSE)
estimation, the second one is based on lineofsight (LOS) part. The respective
analytical expressions of uplink data rate are given for these two methods. Due
to the existence of pilot contamination, the uplink data rate of pilotassisted
LMMSE estimation method approaches to a finite value (we name it as asymptotic
rate in the paper) when the BS antenna number is very large. However, the data
rate of LOS method goes linearly with the number of BS antennas. The expression
of the uplink rate of LOS method also show that for Ricean channel, the spacial
correlation between the BS antennas may not only decrease the rate, but also
increase the rate, which depends on the locations of the users. This conclusion
explains why the spacial correlation may increase, rather than decrease the
data rate of pilotassisted LMMSE. We also discuss the power scaling law of the
two methods, and the asymptotic expressions of the two methods are the same and
both independent of the antenna correlation.

We consider the problem of learning highdimensional Gaussian graphical
models. The graphical lasso is one of the most popular methods for estimating
Gaussian graphical models. However, it does not achieve the oracle rate of
convergence. In this paper, we propose the graphical nonconvex optimization for
optimal estimation in Gaussian graphical models, which is then approximated by
a sequence of convex programs. Our proposal is computationally tractable and
produces an estimator that achieves the oracle rate of convergence. The
statistical error introduced by the sequential approximation using the convex
programs are clearly demonstrated via a contraction property. The rate of
convergence can be further improved using the notion of sparsity pattern. The
proposed methodology is then extended to semiparametric graphical models. We
show through numerical studies that the proposed estimator outperforms other
popular methods for estimating Gaussian graphical models.

We propose a computational framework named iterative local adaptive
majorizeminimization (ILAMM) to simultaneously control algorithmic complexity
and statistical error when fitting high dimensional models. ILAMM is a
twostage algorithmic implementation of the local linear approximation to a
family of folded concave penalized quasilikelihood. The first stage solves a
convex program with a crude precision tolerance to obtain a coarse initial
estimator, which is further refined in the second stage by iteratively solving
a sequence of convex programs with smaller precision tolerances. Theoretically,
we establish a phase transition: the first stage has a sublinear iteration
complexity, while the second stage achieves an improved linear rate of
convergence. Though this framework is completely algorithmic, it provides
solutions with optimal statistical performances and controlled algorithmic
complexity for a large family of nonconvex optimization problems. The iteration
effects on statistical errors are clearly demonstrated via a contraction
property. Our theory relies on a localized version of the sparse/restricted
eigenvalue condition, which allows us to analyze a large family of loss and
penalty functions and provide optimality guarantees under very weak assumptions
(For example, ILAMM requires much weaker minimal signal strength than other
procedures). Thorough numerical results are provided to support the obtained
theory.

The strength of hydrogen bonding in water is stronger than that of van der
waals interaction, therefore water may play an important role in the process of
hydrophobic effects. When a hydrophobic solute is dissolved into water, an
interface appears between the solute and water. To understand the mechanism of
hydrophobic effects, it is necessary to study the structure of water and
solute/water interface. In this study, based on the structural studies on water
and air/water interface, the hydration free energy is derived, and utilized to
investigate the physical origin of hydrophobic effects. According to the
discussion on hydration free energy, with increasing solute size, it can be
divided into the initial and hydrophobic solvation processes, respectively. In
the initial solvation process, hydration free energy is dominated by the
hydrogen bonding in interfacial water (topmost water layer at the solute/water
interface). However, in the hydrophobic solvation process, the hydration free
energy is related to hydrogen bondings of both bulk water and interfacial
water. Additionally, various dissolved behaviors of solutes can be expected for
different solvation processes. From this, hydrophobic effects can be ascribed
to the competition between the hydrogen bondings in bulk water and those in
interfacial water.

For complexity of the heterogeneous minimum spanning forest problem has not
been determined, we reduce 3SAT which is NPcomplete to 2heterogeneous
minimum spanning forest problem to prove this problem is NPhard and spread
result to general problem, which determines complexity of this problem. It
provides a theoretical basis for the future designing of approximation
algorithms for the problem.

The finite element analysis of high frequency vibrations of quartz crystal
plates is a necessary process required in the design of quartz crystal
resonators of precision types for applications in filters and sensors. The
anisotropic materials and extremely high frequency in radiofrequency range of
resonators determine that vibration frequency spectra are complicated with
strong couplings of large number of different vibration modes representing
deformations which do not appear in usual structural problems. For instance,
the higherorder thicknessshear vibrations usually representing the sharp
deformation of thin plates in the thickness direction, expecting the analysis
is to be done with refined meshing schemes along the relatively small thickness
and consequently the large plane area. To be able to represent the precise
vibration mode shapes, a very large number of elements are needed in the finite
element analysis with either the threedimensional theory or the higherorder
plate theory, although considerable reduction of numbers of degreeoffreedom
(DOF) are expected for the twodimensional analysis without scarifying the
accuracy. In this paper, we reviewed the software architecture for the analysis
and demonstrated the evaluation and tuning of parameters for the improvement of
the analysis with problems of elements with a large number of DOF in each node,
or a problem with unusually large bandwidth of the banded stiffness and mass
matrices in comparison with conventional finite element formulation. Such a
problem can be used as an example for the optimization and tuning of problems
from multiphysics analysis which are increasingly important in applications
with excessive large number of DOF and bandwidth in engineering.

As an extension of the FourColor Theorem it is conjectured that every planar
graph of oddgirth at least $2k+1$ admits a homomorphism to
$PC_{2k}=(\mathbb{Z}_2^{2k}, \{e_1, e_2, ...,e_{2k}, J\})$ where $e_i$'s are
standard basis and $J$ is all 1 vector. Noting that $PC_{2k}$ itself is of
oddgirth $2k+1$, in this work we show that if the conjecture is true, then
$PC_{2k}$ is an optimal such a graph both with respect to number of vertices
and number of edges. The result is obtained using the notion of walkpower of
graphs and their clique numbers.
An analogous result is proved for bipartite signed planar graphs of
unbalancedgirth $2k$. The work is presented on a uniform frame work of planar
consistent signed graphs.

First principles calculations based on density functional theory reveal some
unusual properties of BN sheet functionalized with hydrogen and fluorine. These
properties differ from those of similarly functionalized graphene even though
both share the same honeycomb structure. (1) Unlike graphene which undergoes a
metal to insulator transition when fully hydrogenated, the band gap of the BN
sheet significantly narrows when fully saturated with hydrogen. Furthermore,
the band gap of the BN sheet can be tuned from 4.7 eV to 0.6 eV and the system
can be a direct or an indirect semiconductor or even a halfmetal depending
upon surface coverage. (2) Unlike graphene, BN sheet has heteroatomic
composition, when codecorated with H and F, it can lead to anisotropic
structures with rich electronic and magnetic properties. (3) Unlike graphene,
BN sheets can be made ferromagnetic, antiferromagnetic, or magnetically
degenerate depending upon how the surface is functionalized. (4) The stability
of magnetic coupling of functionalized BN sheet can be further modulated by
applying external strain. Our study highlights the potential of functionalized
BN sheets for novel applications.

The structural, vibrational, energetic and electronic properties of hydrogen
at the stoichiometric RuO2(110) termination are studied using density
functional theory. The oxide surface is found to stabilize both molecular and
dissociated H2. The most stable configuration in form of hydroxyl groups
(monohydrides) at the undercoordinated O^br surface anions is at low
temperatures accessed via a molecular state at the undercoordinated Ru^cus
atoms (dihydrogen) and a second precursor in form of a waterlike species
(dihydride) at the O^br sites. This complex picture of the lowtemperature
dissociation kinetics of H2 at RuO2(110) is in agreement with existing data
from highresolution energyloss spectroscopy and temperature programmed
desorption. Hydrogen adsorption at O^br sites increases the reactivity of the
neighboring Ru^cus sites, which are believed to be the active sites in
catalytic oxidation reactions.

Combining densityfunctional theory and thermodynamics we compute the phase
diagram of surface structures of rutile RuO2 (110) in equilibrium with water
vapor in the complete range of experimentally accessible gas phase conditions.
Through the formation of hydroxyl or waterlike groups, already lowest
concentrations of hydrogen in the gas phase are sufficient to stabilize an
oxygenrich polar oxide termination even at very low oxygen pressure.