arXiv subjects

Lu Zhang

Publications and source records attributed to Lu Zhang.

At least 469 records · Page 26Linked to original sources

arXiv2016

Achieving non-discrimination in data release

Discrimination discovery and prevention/removal are increasingly important tasks in data mining. Discrimination discovery aims to unveil discriminatory practices on the protected attribute (e.g., gender) by analyzing the dataset of historical decision records, and discrimination prevention aims to remove discrimination by modifying the biased data before conducting predictive analysis. In this paper, we show that the key to discrimination discovery and prevention is to find the meaningful partitions that can be used to provide quantitative evidences for the judgment of discrimination. With the support of the causal graph, we present a graphical condition for identifying a meaningful partition. Based on that, we develop a simple criterion for the claim of non-discrimination, and propose discrimination removal algorithms which accurately remove discrimination while retaining good data utility. Experiments using real datasets show the effectiveness of our approaches.

cs.LG↗

arXiv2016

A causal framework for discovering and removing direct and indirect discrimination

Anti-discrimination is an increasingly important task in data science. In this paper, we investigate the problem of discovering both direct and indirect discrimination from the historical data, and removing the discriminatory effects before the data is used for predictive analysis (e.g., building classifiers). We make use of the causal network to capture the causal structure of the data. Then we model direct and indirect discrimination as the path-specific effects, which explicitly distinguish the two types of discrimination as the causal effects transmitted along different paths in the network. Based on that, we propose an effective algorithm for discovering direct and indirect discrimination, as well as an algorithm for precisely removing both types of discrimination while retaining good data utility. Different from previous works, our approaches can ensure that the predictive models built from the modified data will not incur discrimination in decision making. Experiments using real datasets show the effectiveness of our approaches.

cs.LG↗

arXiv2016

How Transferable are Neural Networks in NLP Applications?

Transfer learning is aimed to make use of valuable knowledge in a source domain to help model performance in a target domain. It is particularly important to neural networks, which are very likely to be overfitting. In some fields like image processing, many studies have shown the effectiveness of neural network-based transfer learning. For neural NLP, however, existing studies have only casually applied transfer learning, and conclusions are inconsistent. In this paper, we conduct systematic case studies and provide an illuminating picture on the transferability of neural networks in NLP.

cs.CL↗

arXiv2016

Sequence to Backward and Forward Sequences: A Content-Introducing Approach to Generative Short-Text Conversation

Using neural networks to generate replies in human-computer dialogue systems is attracting increasing attention over the past few years. However, the performance is not satisfactory: the neural network tends to generate safe, universally relevant replies which carry little meaning. In this paper, we propose a content-introducing approach to neural network-based generative dialogue systems. We first use pointwise mutual information (PMI) to predict a noun as a keyword, reflecting the main gist of the reply. We then propose seq2BF, a "sequence to backward and forward sequences" model, which generates a reply containing the given keyword. Experimental results show that our approach significantly outperforms traditional sequence-to-sequence models in terms of human evaluation and the entropy measure, and that the predicted keyword can appear at an appropriate position in the reply.

cs.CL↗

arXiv2016

A New Method for Computing $φ$-functions and Their Condition Numbers of Large Sparse Matrices

We propose a new method for computing the $φ$-functions of large sparse matrices with low rank or fast decaying singular values. The key is to reduce the computation of $φ_{\ell}$-functions of a large matrix to $φ_{\ell+1}$-functions of some $r$-by-$r$ matrices, where $r$ is the numerical rank of the large matrix in question. Some error analysis on the new method is given. Furthermore, we propose two novel strategies for estimating 2-norm condition numbers of the $φ$-functions. Numerical experiments illustrate the numerical behavior of the new algorithms and show the effectiveness of our theoretical results.

math.NA↗

arXiv2016

Distilling Word Embeddings: An Encoding Approach

Distilling knowledge from a well-trained cumbersome network to a small one has recently become a new research topic, as lightweight neural networks with high performance are particularly in need in various resource-restricted systems. This paper addresses the problem of distilling word embeddings for NLP tasks. We propose an encoding approach to distill task-specific knowledge from a set of high-dimensional embeddings, which can reduce model complexity by a large margin as well as retain high accuracy, showing a good compromise between efficiency and performance. Experiments in two tasks reveal the phenomenon that distilling knowledge from cumbersome embeddings is better than directly training neural networks with small embeddings.

cs.CL↗

arXiv2016

Application of Origen2.1 in the decay photon spectrum calculation of spallation products

Origen2.1 is a widely used computer code for calculating the burnup, decay, and processing of radioactive materials. However, the nuclide library of Origen2.1 is used for existing reactors like pressurized water reactor, to calculate the photon spectrum released by the decay of spallation products, we have made specific libraries for the ADS tungsten spallation target, based on the results given by a Monte Carlo code: FLUKA. All the data used to make the Origen2.1 libraries is obtained from Nuclear structure & decay Data (NuDat2.6). The accumulated activity of spallation products and the contribution of nuclides to photon emission are given in this paper.

nucl-th↗

arXiv2016

Resonance of Gaussian electromagnetic field to the high frequency gravitational waves

We consider a Gaussian Beam (GB) resonant system for high frequency gravitational waves (HFGWs) detection. At present, we find the optimal signal strength in theory through setting the magnetic component of GB in a standard gaussian form. Under the synchro-resonance condition, we study the signal strength (i.e., transverse perturbative photon fluxes) from the relic HFGWs (predicted by ordinary inflationary model) and the braneworld HFGWs (from braneworld scenarios). Both of them would generate potentially detectable transverse perturbative photon fluxes (PPFs). Furthermore we find optimal system parameters and the relationship between frequency and effective width of energy fluxes accumulation.

gr-qc↗

arXiv2016

Natural Language Inference by Tree-Based Convolution and Heuristic Matching

In this paper, we propose the TBCNN-pair model to recognize entailment and contradiction between two sentences. In our model, a tree-based convolutional neural network (TBCNN) captures sentence-level semantics; then heuristic matching layers like concatenation, element-wise product/difference combine the information in individual sentences. Experimental results show that our model outperforms existing sentence encoding-based approaches by a large margin.

cs.CL↗

arXiv2016

A novel three-axis cylindrical hohlraum designed for inertial confinement fusion ignition

A novel ignition hohlraum for indirect-drive inertial confinement fusion is proposed, which is named as three-axis cylindrical hohlraum (TACH). TACH is a kind of 6 laser entrance holes (LEHs) hohlraum, which is made of three cylindrical hohlraums orthogonally jointed. Laser beams are injected through every entrance hole with the same incident angle of 55°. The view-factor simulation result shows that the time-varying drive asymmetry of TACH is no more than 1.0% in the whole drive pulse period without any supplementary technology such as beam phasing etc. Its coupling efficiency of TACH is close to that of 6 LEHs spherical hohlraum with corresponding size. Its plasma-filling time is close to typical cylindrical ignition hohlraum. Its laser plasma interaction has as low backscattering as the outer cone of the cylindrical ignition hohlraum. Therefore, the proposed hohlraum provides a competitive candidate for ignition hohlraum.

physics.plasm-ph↗

arXiv2016

Power Side Channels in Security ICs: Hardware Countermeasures

Power side-channel attacks are a very effective cryptanalysis technique that can infer secret keys of security ICs by monitoring the power consumption. Since the emergence of practical attacks in the late 90s, they have been a major threat to many cryptographic-equipped devices including smart cards, encrypted FPGA designs, and mobile phones. Designers and manufacturers of cryptographic devices have in response developed various countermeasures for protection. Attacking methods have also evolved to counteract resistant implementations. This paper reviews foundational power analysis attack techniques and examines a variety of hardware design mitigations. The aim is to highlight exposed vulnerabilities in hardware-based countermeasures for future more secure implementations.

cs.CR↗

arXiv2016

Universal Optimal Estimation of the Polarization of Light with Arbitrary Photon Statistics

A universal and optimal method for the polarimetry of light with arbitrary photon statistics is presented. The method is based on the continuous maximum-likelihood positive operator-valued measure (ML-POVM) for pure polarization states over the surface of the Bloch sphere. The success probability and the mean fidelity are used as the figures of merit to show its performance. The POVM is found to attain the collective bound of polarization estimation with respect to the mean fidelity. As demonstrations, explicit results for the N photon Fock state, the phase-randomized coherent state (Poisson distribution), and the thermal light are obtained. It is found that the estimation performances for the Fock state and the Poisson distribution are almost identical, while that for the thermal light is much worse. This suggests that thermal light leaks less information to an eavesdropper and hence could potentially provide more security in polarization-encoded quantum communication protocols than a single-mode laser beam as customarily considered. Finally, comparisons against an optimal adaptive measurement with classical communications are made to show the better and more stable performance of the continuous ML-POVM.

quant-ph↗

arXiv2016

Superconducting nanowire single photon detector at 532 nm and demonstration in satellite laser ranging

Superconducting nanowire single-photon detectors (SNSPDs) at a wavelength of 532 nm were designed and fabricated aiming to satellite laser ranging (SLR) applications. The NbN SNSPDs were fabricated on one-dimensional photonic crystals with a sensitive-area diameter of 42 um. The devices were coupled with multimode fiber (phi=50um) and exhibited a maximum system detection efficiency of 75% at an extremely low dark count rate of <0.1 Hz. An SLR experiment using an SNSPD at a wavelength of 532 nm was successfully demonstrated. The results showed a depth ranging with a precision of ~8.0 mm for the target satellite LARES, which is ~3,000 km away from the ground ranging station at the Sheshan Observatory.

cond-mat.supr-con↗

arXiv2016

Backward and Forward Language Modeling for Constrained Sentence Generation

Recent language models, especially those based on recurrent neural networks (RNNs), make it possible to generate natural language from a learned probability. Language generation has wide applications including machine translation, summarization, question answering, conversation systems, etc. Existing methods typically learn a joint probability of words conditioned on additional information, which is (either statically or dynamically) fed to RNN's hidden layer. In many applications, we are likely to impose hard constraints on the generated texts, i.e., a particular word must appear in the sentence. Unfortunately, existing approaches could not solve this problem. In this paper, we propose a novel backward and forward language model. Provided a specific word, we use RNNs to generate previous words and future words, either simultaneously or asynchronously, resulting in two model variants. In this way, the given word could appear at any position in the sentence. Experimental results show that the generated texts are comparable to sequential LMs in quality.

cs.CL↗

arXiv2015

Convolutional Neural Networks over Tree Structures for Programming Language Processing

Programming language processing (similar to natural language processing) is a hot research topic in the field of software engineering; it has also aroused growing interest in the artificial intelligence community. However, different from a natural language sentence, a program contains rich, explicit, and complicated structural information. Hence, traditional NLP models may be inappropriate for programs. In this paper, we propose a novel tree-based convolutional neural network (TBCNN) for programming language processing, in which a convolution kernel is designed over programs' abstract syntax trees to capture structural information. TBCNN is a generic architecture for programming language processing; our experiments show its effectiveness in two different program analysis tasks: classifying programs according to functionality, and detecting code snippets of certain patterns. TBCNN outperforms baseline methods, including several neural models for NLP.

cs.LG↗

arXiv2015

Dynamical phases in a one-dimensional chain of Heterospecies Rydberg atoms with next-nearest neighbor interactions

We theoretically investigate the dynamical phase diagram of a one-dimensional chain of laser-excited two-species Rydberg atoms. The existence of a variety of unique dynamical phases in the experimentally-achievable parameter region is predicted under the mean-field approximation, and the change of those phases when the effect of the next-nearest neighbor interaction is included is further discussed. In particular we find the competition of the strong Rydberg-Rydberg interactions and the optical excitation imbalance can lead to the presence of complex multiple chaotic phases, which are highly sensitive to the initial Rydberg-state population and the strength of the next-nearest neighbor interactions.

physics.atom-ph↗

arXiv2015

On End-to-End Program Generation from User Intention by Deep Neural Networks

This paper envisions an end-to-end program generation scenario using recurrent neural networks (RNNs): Users can express their intention in natural language; an RNN then automatically generates corresponding code in a characterby-by-character fashion. We demonstrate its feasibility through a case study and empirical analysis. To fully make such technique useful in practice, we also point out several cross-disciplinary challenges, including modeling user intention, providing datasets, improving model architectures, etc. Although much long-term research shall be addressed in this new field, we believe end-to-end program generation would become a reality in future decades, and we are looking forward to its practice.

cs.SE↗

arXiv2015

$L^p$ Boundedness of rough Bi-parameter Fourier Integral Operators

In this paper, we will investigate the boundedness of the bi-parameter Fourier integral operators (or FIOs for short) of the following form: $$T(f)(x)=\frac{1}{(2π)^{2n}}\int_{\mathbb{R}^{2n}}e^{iφ(x,ξ,η)}\cdot a(x,ξ,η)\cdot\widehat{f}(ξ,η)dξdη,$$ where for $x=(x_1,x_2)\in \mathbb{R}^{n}\times \mathbb{R}^{n}$ and $ξ,η\in \mathbb{R}^{n}\setminus\{0\}$, the amplitude $a(x,ξ,η)\in L^\infty BS^m_ρ$ and the phase function is of the form $ φ(x,ξ,η)=φ_1(x_1,ξ)+φ_2(x_2,η)$ with $\quad φ_1,φ_2 \in L^\infty Φ^2 (\mathbb{R}^{n}\times\mathbb{R}^{n}\setminus\{0\})$ and $φ(x, ξ, η)$ satisfies a certain rough non-degeneracy condition. The study of these operators are motivated by the $L^p$ estimates for one-parameter FIOs and bi-parameter Fourier multipliers and pseudo-differential operators. We will first define the bi-parameter FIOs and then study the $L^p$ boundedness of such operators when their phase functions have compact support in frequency variables with certain necessary non-degeneracy conditions. We will then establish the $L^p$ boundedness of the more general FIOs with amplitude $a(x,ξ,η)\in L^\infty BS^m_ρ$ and non-smooth phase function $φ(x,ξ,η)$ on $x$ satisfying a rough non-degeneracy condition.

math.AP↗