arXiv subjects

Jin Yu

Publications and source records attributed to Jin Yu.

44 records · Page 3Linked to original sources

arXiv2014

Magnetic Proximity Effect and Interlayer Exchange Coupling of Ferromagnetic/Topological Insulator/Ferromagnetic Trilayer

Magnetic proximity effect between topological insulator (TI) and ferromagnetic insulator (FMI) is considered to have great potential in spintronics. However, a complete determination of interfacial magnetic structure has been highly challenging. We theoretically investigate the interlayer exchange coupling of two FMIs separated by a TI thin film, and show that the particular electronic states of the TI contributing to the proximity effect can be directly identified through the coupling behavior between two FMIs, together with a tunability of coupling constant. Such FMI/TI/FMI structure not only serves as a platform to clarify the magnetic structure of FMI/TI interface, but also provides insights into designing the magnetic storage devices with ultrafast response.

cond-mat.mes-hall↗

arXiv2014

Efficient fidelity control by stepwise nucleotide selection in polymerase elongation

Polymerases select nucleotides before incorporating them for chemical synthesis during gene replication or transcription. How the selection proceeds stepwise efficiently to achieve sufficiently high fidelity and speed is essential for polymerase function. We examined step-by-step selections that have conformational transition rates tuned one at time in the polymerase elongation cycle, with a controlled differentiation free energy at each checkpoint. The elongation is sustained at non-equilibrium steady state with constant free energy input and heat dissipation. It is found that error reduction capability does not improve for selection checkpoints down the reaction path. Hence, it is essential to select early to achieve an efficient fidelity control. In particular, for two consecutive selections that reject the wrong substrate back and inhibit it forward from a same kinetic state, the same error rates are obtained at the same free energy differentiation. The initial screening is indispensible for maintaining the elongation speed high, as the wrong nucleotides can be removed quickly and replaced by the right nucleotides at the entry. Overall, the elongation error rate can be repeatedly reduced through multiple selection checkpoints. The study provides a theoretical framework to conduct further detailed researches, and assists engineering and redesign of related enzymes.

physics.bio-ph↗

arXiv2013

Effective Absorption Enhancement in Small Molecule Organic Solar Cells by Employing Trapezoid Gratings

We demonstrate the optical absorption has been enhanced in the small molecule organic solar cells by employing trapezoid grating structure. The enhanced absorption is mainly attributed to both waveguide modes and surface plasmon modes, which has been simulated by using finite-difference time-domain method. The simulated results show that the surface plasmon along the semitransparent metallic Ag anode is excited by introducing the periodical trapezoid gratings, which induce high intensity field increment in the donor layer. Meanwhile, the waveguide modes result a high intensity field in acceptor layer. The increment of field improves the absorption of organic solar cells, significantly, which has been demonstrated by simulating the electrical properties. The simulated results exhibiting 31 % increment of the short-circuit current has been achieved in the optimized device, which is supported by the experimental measurement. The power conversion efficiency of the grating sample obtained in experiment exhibits an enhancement of 7.7 %.

physics.optics↗

arXiv2013

A New Paradigm to Half-Metallicity in Graphene Nanoribbons

In contrast to the well recognized transverse-electric-field-induced half-metallicity in zigzag graphene nanoribbons, here we demonstrate by first-principles calculations that zigzag graphene nanoribbons sandwiched between hexagonal boron nitride nanoribbons or sheets can be tuned into half-metal simply by a bias voltage or a moderate compressive strain. The half-metallicity is attributed to an enhanced coupling effect of spontaneous polarization and asymmetrical exchange correlation along the ribbon width. The findings should open a viable route for efficient spin-resolved band engineering in graphene based devices that are compatible with the current technology of semiconductor industry.

cond-mat.mes-hall↗

arXiv2012

The Entire Quantile Path of a Risk-Agnostic SVM Classifier

A quantile binary classifier uses the rule: Classify x as +1 if P(Y = 1|X = x) >= t, and as -1 otherwise, for a fixed quantile parameter t {[0, 1]. It has been shown that Support Vector Machines (SVMs) in the limit are quantile classifiers with t = 1/2 . In this paper, we show that by using asymmetric cost of misclassification SVMs can be appropriately extended to recover, in the limit, the quantile binary classifier for any t. We then present a principled algorithm to solve the extended SVM classifier for all values of t simultaneously. This has two implications: First, one can recover the entire conditional distribution P(Y = 1|X = x) = t for t {[0, 1]. Second, we can build a risk-agnostic SVM classifier where the cost of misclassification need not be known apriori. Preliminary numerical experiments show the effectiveness of the proposed algorithm.

cs.LG↗

arXiv2011

Incremental Top-k List Comparison Approach to Robust Multi-Structure Model Fitting

Random hypothesis sampling lies at the core of many popular robust fitting techniques such as RANSAC. In this paper, we propose a novel hypothesis sampling scheme based on incremental computation of distances between partial rankings (top-$k$ lists) derived from residual sorting information. Our method simultaneously (1) guides the sampling such that hypotheses corresponding to all true structures can be quickly retrieved and (2) filters the hypotheses such that only a small but very promising subset remain. This permits the usage of simple agglomerative clustering on the surviving hypotheses for accurate model selection. The outcome is a highly efficient multi-structure robust estimation technique. Experiments on synthetic and real data show the superior performance of our approach over previous methods.

cs.CV↗

arXiv2010

A Quasi-Newton Approach to Nonsmooth Convex Optimization Problems in Machine Learning

We extend the well-known BFGS quasi-Newton method and its memory-limited variant LBFGS to the optimization of nonsmooth convex objectives. This is done in a rigorous fashion by generalizing three components of BFGS to subdifferentials: the local quadratic model, the identification of a descent direction, and the Wolfe line search conditions. We prove that under some technical conditions, the resulting subBFGS algorithm is globally convergent in objective function value. We apply its memory-limited variant (subLBFGS) to L_2-regularized risk minimization with the binary hinge loss. To extend our algorithm to the multiclass and multilabel settings, we develop a new, efficient, exact line search algorithm. We prove its worst-case time complexity bounds, and show that our line search can also be used to extend a recently developed bundle method to the multiclass and multilabel settings. We also apply the direction-finding component of our algorithm to L_1-regularized risk minimization with logistic loss. In all these contexts our methods perform comparable to or better than specialized state-of-the-art solvers on a number of publicly available datasets. An open source implementation of our algorithms is freely available.

stat.ML↗

arXiv2009

Exponential Family Graph Matching and Ranking

We present a method for learning max-weight matching predictors in bipartite graphs. The method consists of performing maximum a posteriori estimation in exponential families with sufficient statistics that encode permutations and data features. Although inference is in general hard, we show that for one very relevant application - web page ranking - exact inference is efficient. For general model instances, an appropriate sampler is readily available. Contrary to existing max-margin matching models, our approach is statistically consistent and, in addition, experiments with increasing sample sizes indicate superior improvement over such models. We apply the method to graph matching in computer vision as well as to a standard benchmark dataset for learning web page ranking, in which we obtain state-of-the-art results, in particular improving on max-margin variants. The drawback of this method with respect to max-margin alternatives is its runtime for large graphs, which is comparatively high.

cs.LG↗