Convex relaxation of mixture regression with efficient algorithms

We develop a convex relaxation of maximum a posteriori estimation of a mixture of regression models. Although our relaxation involves a semidefinite matrix variable, we reformulate the problem to eliminate the need for general semidefinite programming. In particular, we provide two reformulations that admit fast algorithms. The first is a max-min spectral reformulation exploiting quasi-Newton descent. The second is a min-min reformulation consisting of fast alternating steps of closed-form updates. We evaluate the methods against Expectation-Maximization in a real problem of motion segmentation from video data.

Minimum maneuver time calculation using convex optimization

The problem of calculating the minimum lap or maneuver time of a nonlinear vehicle, which is linearized at each time step, is formulated as a convex optimization problem. The formulation provides an alternative to previously used quasi-steady-state analysis or nonlinear optimization. Key steps are: the use of model predictive control; expressing the minimum time problem as one of maximizing distance traveled along the track centerline; and linearizing the track and vehicle trajectories by expressing them as small displacements from a fixed reference. A consequence of linearizing the vehicle dynamics is that nonoptimal steering control action can be generated, but attention to the constraints and the cost function minimizes the effect. Optimal control actions and vehicle responses for a 90 deg bend are presented and compared to the nonconvex nonlinear programming solution. Copyright © 2013 by ASME.

Effect of size, composition, and morphology on magnetic performance: First-order reversal curves evaluation of iron oxide nanoparticles

© 2014 AIP Publishing LLC. Superparamagnetic nanoparticles are employed in a broad range of applications that demand detailed magnetic characterization for superior performance, e.g., in drug delivery or cancer treatment. Magnetic hysteresis measurements provide information on saturation magnetization and coercive force for bulk material but can be equivocal for particles having a broad size distribution. Here, first-order reversal curves (FORCs) are used to evaluate the effective magnetic particle size and interaction between equally sized magnetic iron oxide (Fe2O3) nanoparticles with three different morphologies: (i) pure Fe2O3, (ii) Janus-like, and (iii) core/shell Fe2O3/SiO2synthesized using flame technology. By characterizing the distribution in coercive force and interaction field from the FORC diagrams, we find that the presence of SiO2in the core/shell structures significantly reduces the average coercive force in comparison to the Janus-like Fe2O3/SiO2and pure Fe2O3particles. This is attributed to the reduction in the dipolar interaction between particles, which in turn reduces the effective magnetic particle size. Hence, FORC analysis allows for a finer distinction between equally sized Fe2O3particles with similar magnetic hysteresis curves that can significantly influence the final nanoparticle performance.

Dynamic P-I and P-V Curves for Semiconductor Lasers and Modulators

In this paper, we propose the dynamic P-V curve for modulator and P-I curve for laser diode, and present a simple approach to deriving the curves from the small-signal frequency responses measured using a microwave network analyzer. The linear response range, modulation efficiency, optimal driving conditions at different frequency can, therefore, be determined. It is demonstrated that the large-signal performance of electro-absorption (EA) modulator and the directly modulated semiconductor lasers can be predicted from the dynamic curved surface. Experiments show a good agreement between the evaluated characteristics and the measured large-signal performance.

Nonradiative recombination effect on photoluminescence decay dynamics in GaInNAs/GaAs quantum wells

The influence of nonradiative recombination on the photoluminescence (PL) decay dynamics in GaInNAs/GaAs quantum wells is studied by time-resolved photoluminescence under various excitation intensities. It is found that the PL decay process strongly depends on the excitation intensity. In particular, under the moderate excitation levels the PL decay curves exhibit unusual nonexponential behavior and show a convex shape. By introducing a new parameter of the effective concentration of nonradiative recombination centers into a rate equation, the observed results are well simulated. The cw PL data further demonstrate the nonradiative recombination effect on the optical properties of GaInNAs/GaAs quantum wells. (c) 2006 American Institute of Physics.

Nonradiative recombination effect on photoluminescence decay dynamics in GaInNAs/GaAs quantum wells - art. no. 61180Z

The nonradiative recombination effect on the photoluminescence (PL) decay dynamics in GaInNAs/GaAs quantum wells is studied by photoluminescence and time-resolved photoluminescence under various excitation intensities and temperatures. It is found that the PL decay dynamics strongly depends on the excitation intensity. In particular, under the moderate excitation levels the PL decay curves exhibit unusual non-exponential behavior and show a convex shape. By introducing a new concept of the effective concentration of nonradiative recombination centers into a rate equation, the observed results are well simulated. In the cw PL measurement, a rapid PL quenching is observed even at very low temperature and is of the excitation power dependence. These results further demonstrate that the non-radiative recombination process plays a very important role on the optical properties of GaInNAs/GaAs quantum wells.

on a class of pseudorandom sequences from elliptic curves over finite fields

Following the idea of Xing et al., we investigate a general method for constructing families of pseudorandom sequences with low correlation and large linear complexity from elliptic curves over finite fields in this correspondence. With the help of the tool of exponential sums on elliptic curves, we study their periods, linear complexities, linear complexity profiles, distributions of r-patterns, periodic correlation, partial period distributions, and aperiodic correlation in detail. The results show that they have nice randomness.

Response amplitude and hydrodynamic force for a buoy over a convex

A buoy as an offshore structure is often placed over a convex such as a caisson or a submerged island. The hydrodynamic fluid/solid interaction becomes more complex due to the convex compared with that on the flat. Both the buoy and the convex are idealized as vertical cylinders. Linear potential theory is used to investigate the response amplitude and the hydrodynamic force for a buoy over a convex due to diffraction and radiation in water of finite depth. These are derived from the total velocity potential. A set of theoretical added mass, damping coefficient, and exciting force expressions have been proposed. Analytical results of the response amplitude and hydrodynamic force are given. Finally, the numerical results show that the effect of the convex on the response amplitude and hydrodynamic force for the buoy is ignored if the size of the convex is relatively smaller.

Reversed capillary electrochromatography: Investigation of peak symmetry for eluting curves of solutes

The transport processes of components in capillary electrochromatographic column was investigated based on the basic model of relaxation theory. A principal transport equation of chromatographic relaxation theory was established and mathematical expressions for eluting curves were obtained under the situations of both capillary electrophoresis and chromatography. Characteristics of peak symmetry and its effecting factors are discussed. Tailing peaks, symmetrical peaks and fronting peaks would be observed simultaneously, which was further proved with reversed capillary electrochromatographic experiments.