arXiv subjects

Tian Lan

Publications and source records attributed to Tian Lan.

At least 217 records · Page 12Linked to original sources

arXiv2021

MILA: Multi-Task Learning from Videos via Efficient Inter-Frame Attention

Prior work in multi-task learning has mainly focused on predictions on a single image. In this work, we present a new approach for multi-task learning from videos via efficient inter-frame local attention (MILA). Our approach contains a novel inter-frame attention module which allows learning of task-specific attention across frames. We embed the attention module in a ``slow-fast'' architecture, where the slower network runs on sparsely sampled keyframes and the light-weight shallow network runs on non-keyframes at a high frame rate. We also propose an effective adversarial learning strategy to encourage the slow and fast network to learn similar features. Our approach ensures low-latency multi-task learning while maintaining high quality predictions. Experiments show competitive accuracy compared to state-of-the-art on two multi-task learning benchmarks while reducing the number of floating point operations (FLOPs) by up to 70\%. In addition, our attention based feature propagation method (ILA) outperforms prior work in terms of task accuracy while also reducing up to 90\% of FLOPs.

cs.CV↗

arXiv2021

WarpDrive: Extremely Fast End-to-End Deep Multi-Agent Reinforcement Learning on a GPU

Deep reinforcement learning (RL) is a powerful framework to train decision-making models in complex environments. However, RL can be slow as it requires repeated interaction with a simulation of the environment. In particular, there are key system engineering bottlenecks when using RL in complex environments that feature multiple agents with high-dimensional state, observation, or action spaces. We present WarpDrive, a flexible, lightweight, and easy-to-use open-source RL framework that implements end-to-end deep multi-agent RL on a single GPU (Graphics Processing Unit), built on PyCUDA and PyTorch. Using the extreme parallelization capability of GPUs, WarpDrive enables orders-of-magnitude faster RL compared to common implementations that blend CPU simulations and GPU models. Our design runs simulations and the agents in each simulation in parallel. It eliminates data copying between CPU and GPU. It also uses a single simulation data store on the GPU that is safely updated in-place. WarpDrive provides a lightweight Python interface and flexible environment wrappers that are easy to use and extend. Together, this allows the user to easily run thousands of concurrent multi-agent simulations and train on extremely large batches of experience. Through extensive experiments, we verify that WarpDrive provides high-throughput and scales almost linearly to many agents and parallel environments. For example, WarpDrive yields 2.9 million environment steps/second with 2000 environments and 1000 agents (at least 100x higher throughput compared to a CPU implementation) in a benchmark Tag simulation. As such, WarpDrive is a fast and extensible multi-agent RL platform to significantly accelerate research and development.

cs.LG↗

arXiv2021

MPD: Moving Target Defense through Communication Protocol Dialects

Communication protocol security is among the most significant challenges of the Internet of Things (IoT) due to the wide variety of hardware and software technologies involved. Moving target defense (MTD) has been adopted as an innovative strategy to solve this problem by dynamically changing target system properties and configurations to obfuscate the attack surface. Nevertheless, the existing work of MTD primarily focuses on lower-level properties (e.g., IP addresses or port numbers), and only a limited number of variations can be generated based on these properties. In this paper, we propose a new approach of MTD through communication protocol dialects (MPD) - which dynamically customizes a communication protocol into various protocol dialects and leverages them to create a moving target defense. Specifically, MPD harnesses a dialect generating function to create protocol dialects and then a mapping function to select one specific dialect for each packet during communication. To keep different network entities in synchronization, we also design a self-synchronization mechanism utilizing a pseudo-random number generator with the input of a pre-shared secret key and previously sent packets. We implement a prototype of MPD and evaluate its feasibility on standard network protocol (i.e., File Transfer Protocol) and internet of things protocol (i.e., Message Queuing Telemetry Transport). The results indicate that MPD can create a moving target defense with protocol dialects to effectively address various attacks - including the denial of service attack and malicious packet modifications - with negligible overhead.

cs.CR↗

arXiv2021

Merlion: A Machine Learning Library for Time Series

We introduce Merlion, an open-source machine learning library for time series. It features a unified interface for many commonly used models and datasets for anomaly detection and forecasting on both univariate and multivariate time series, along with standard pre/post-processing layers. It has several modules to improve ease-of-use, including visualization, anomaly score calibration to improve interpetability, AutoML for hyperparameter tuning and model selection, and model ensembling. Merlion also provides a unique evaluation framework that simulates the live deployment and re-training of a model in production. This library aims to provide engineers and researchers a one-stop solution to rapidly develop models for their specific time series needs and benchmark them across multiple time series datasets. In this technical report, we highlight Merlion's architecture and major functionalities, and we report benchmark numbers across different baseline models and ensembles.

cs.LG↗

arXiv2021

DQ-SGD: Dynamic Quantization in SGD for Communication-Efficient Distributed Learning

Gradient quantization is an emerging technique in reducing communication costs in distributed learning. Existing gradient quantization algorithms often rely on engineering heuristics or empirical observations, lacking a systematic approach to dynamically quantize gradients. This paper addresses this issue by proposing a novel dynamically quantized SGD (DQ-SGD) framework, enabling us to dynamically adjust the quantization scheme for each gradient descent step by exploring the trade-off between communication cost and convergence error. We derive an upper bound, tight in some cases, of the convergence error for a restricted family of quantization schemes and loss functions. We design our DQ-SGD algorithm via minimizing the communication cost under the convergence error constraints. Finally, through extensive experiments on large-scale natural language processing and computer vision tasks on AG-News, CIFAR-10, and CIFAR-100 datasets, we demonstrate that our quantization scheme achieves better tradeoffs between the communication cost and learning performance than other state-of-the-art gradient quantization methods.

cs.LG↗

arXiv2021

Communication Efficient Federated Learning with Adaptive Quantization

Federated learning (FL) has attracted tremendous attentions in recent years due to its privacy preserving measures and great potentials in some distributed but privacy-sensitive applications like finance and health. However, high communication overloads for transmitting high-dimensional networks and extra security masks remains a bottleneck of FL. This paper proposes a communication-efficient FL framework with Adaptive Quantized Gradient (AQG) which adaptively adjusts the quantization level based on local gradient's update to fully utilize the heterogeneousness of local data distribution for reducing unnecessary transmissions. Besides, the client dropout issues are taken into account and the Augmented AQG is developed, which could limit the dropout noise with an appropriate amplification mechanism for transmitted gradients. Theoretical analysis and experiment results show that the proposed AQG leads to 25%-50% of additional transmission reduction as compared to existing popular methods including Quantized Gradient Descent (QGD) and Lazily Aggregated Quantized (LAQ) gradient-based method without deteriorating convergence properties. Particularly, experiments with heterogenous data distributions corroborate a more significant transmission reduction compared with independent identical data distributions. Meanwhile, the proposed AQG is robust to a client dropping rate up to 90% empirically, and the Augmented AQG manages to further improve the FL system's communication efficiency with the presence of moderate-scale client dropouts commonly seen in practical FL scenarios.

cs.DC↗

arXiv2021

On the Approximability of Related Machine Scheduling under Arbitrary Precedence

Distributed computing systems often need to consider the scheduling problem involving a collection of highly dependent data-processing tasks that must work in concert to achieve mission-critical objectives. This paper considers the unrelated machine scheduling problem for minimizing weighted sum completion time under arbitrary precedence constraints and on heterogeneous machines with different processing speeds. The problem is known to be strongly NP-hard even in the single machine setting. By making use of Queyranne's constraint set and constructing a novel Linear Programming relaxation for the scheduling problem under arbitrary precedence constraints, our results in this paper advance the state of the art. We develop a $2(1+(m-1)/D)$-approximation algorithm (and $2(1+(m-1)/D)+1$-approximation) for the scheduling problem with zero release time (and arbitrary release time), where $m$ is the number of servers and $D$ is the task-skewness product. The algorithm can be efficiently computed in polynomial time using the Ellipsoid method and achieves nearly optimal performance in practice as $D>O(m)$ when the number of tasks per job to schedule is sufficiently larger than the number of machines available. Our implementation and evaluation using a heterogeneous testbed and real-world benchmarks confirms significant improvement in weighted sum completion time for dependent computing tasks.

cs.DC↗

arXiv2021

Integrated Reasoning Engine for Pointer-related Code Clone Detection

Detecting similar code fragments, usually referred to as code clones, is an important task. In particular, code clone detection can have significant uses in the context of vulnerability discovery, refactoring and plagiarism detection. However, false positives are inevitable and always require manual reviews. In this paper, we propose Twin-Finder+, a novel closed-loop approach for pointer-related code clone detection that integrates machine learning and symbolic execution techniques to achieve precision. Twin-Finder+ introduces a formal verification mechanism to automate such manual reviews process. Our experimental results show Twin-Finder+ that can remove 91.69% false positives in average. We further conduct security analysis for memory safety using real-world applications, Links version 2.14 and libreOffice-6.0.0.1. Twin-Finder+ is able to find 6 unreported bugs in Links version 2.14 and one public patched bug in libreOffice-6.0.0.1.

cs.SE↗

arXiv2021

Sobolev Orthogonal Polynomials on the Sierpinski Gasket

We develop a theory of Sobolev orthogonal polynomials on the Sierpiński gasket ($SG$). These orthogonal polynomials arise through the Gram-Schmidt orthogonalisation process applied on the set of monomials on $SG$ using several notions of a Sobolev inner products. After establishing some recurrence relations for these orthogonal polynomials, we give estimates for their $L^2$, $L^\infty$ and Sobolev norms, and study their asymptotic behaviour. Finally, we study the properties of zero sets of polynomials and develop fast computational tools to explore applications to quadrature and interpolation.

math.CA↗

arXiv2020

Self-attention Comparison Module for Boosting Performance on Retrieval-based Open-Domain Dialog Systems

Since the pre-trained language models are widely used, retrieval-based open-domain dialog systems, have attracted considerable attention from researchers recently. Most of the previous works select a suitable response only according to the matching degree between the query and each individual candidate response. Although good performance has been achieved, these recent works ignore the comparison among the candidate responses, which could provide rich information for selecting the most appropriate response. Intuitively, better decisions could be made when the models can get access to the comparison information among all the candidate responses. In order to leverage the comparison information among the candidate responses, in this paper, we propose a novel and plug-in Self-attention Comparison Module for retrieval-based open-domain dialog systems, called SCM. Extensive experiment results demonstrate that our proposed self-attention comparison module effectively boosts the performance of the existing retrieval-based open-domain dialog systems. Besides, we have publicly released our source codes for future research.

cs.CL↗

arXiv2020

Ultra-Fast, Low-Storage, Highly Effective Coarse-grained Selection in Retrieval-based Chatbot by Using Deep Semantic Hashing

We study the coarse-grained selection module in retrieval-based chatbot. Coarse-grained selection is a basic module in a retrieval-based chatbot, which constructs a rough candidate set from the whole database to speed up the interaction with customers. So far, there are two kinds of approaches for coarse-grained selection module: (1) sparse representation; (2) dense representation. To the best of our knowledge, there is no systematic comparison between these two approaches in retrieval-based chatbots, and which kind of method is better in real scenarios is still an open question. In this paper, we first systematically compare these two methods from four aspects: (1) effectiveness; (2) index stoarge; (3) search time cost; (4) human evaluation. Extensive experiment results demonstrate that dense representation method significantly outperforms the sparse representation, but costs more time and storage occupation. In order to overcome these fatal weaknesses of dense representation method, we propose an ultra-fast, low-storage, and highly effective Deep Semantic Hashing Coarse-grained selection method, called DSHC model. Specifically, in our proposed DSHC model, a hashing optimizing module that consists of two autoencoder models is stacked on a trained dense representation model, and three loss functions are designed to optimize it. The hash codes provided by hashing optimizing module effectively preserve the rich semantic and similarity information in dense vectors. Extensive experiment results prove that, our proposed DSHC model can achieve much faster speed and lower storage than sparse representation, with limited performance loss compared with dense representation. Besides, our source codes have been publicly released for future research.

cs.CL↗

arXiv2020

LG/CY correspondence between $tt^*$ geometries

The concept of $tt^*$ geometric structure was introduced by physicists (see \cite{CV1, BCOV} and references therein) , and then studied firstly in mathematics by C. Hertling \cite{Het1}. It is believed that the $tt^*$ geometric structure contains the whole genus $0$ information of a two dimensional topological field theory. In this paper, we propose the LG/CY correspondence conjecture for $tt^*$ geometry and obtain the following result. Let $f\in\mathbb{C}[z_0, \dots, z_{n+2}]$ be a nondegenerate homogeneous polynomialof degree $n+2$, then it defines a Calabi-Yau model represented by a Calabi-Yau hypersurface $X_f$ in $\mathbb{CP}^{n+1}$ or a Landau-Ginzburg model represented by a hypersurface singularity $(\mathbb{C}^{n+2}, f)$, both can be written as a $tt^*$ structure. We proved that there exists a $tt^*$ substructure on Landau-Ginzburg side, which should correspond to the $tt^*$ structure from variation of Hodge structures in Calabi-Yau side. We build the isomorphism of almost all structures in $tt^*$ geometries between these two models except the isomorphism between real structures.

math.AG↗

arXiv2020

Search-free DOA Estimation Method Based on Tensor Decomposition and Polynomial Rooting for Transmit Beamspace MIMO Radar

In order to improve the accuracy and resolution for transmit beamspace multiple-input multiple-output (MIMO) radar, a search-free direction-of-arrival (DOA) estimation method based on tensor decomposition and polynomial rooting is proposed. In the proposed method, a 3-order tensor is firstly designed to model the received signal of MIMO radar on the basis of the multi-linear property. Then, the factor matrix with target DOA information is obtained by the tensor decomposition via alternating least squares (ALS) algorithm, and subsequently the DOA estimation is converted into the independent minimization problem. By exploiting the Vandermonde structure of the transmit steering vector, a polynomial function is constructed to solve the minimization problem via polynomial rooting. The factor matrix contained in the coefficients of the polynomial can be regarded as a block matrix in the generalized sidelobe canceller (GSC), which accordingly forms a unique deep null in the direction of target in the transmit beampattern. The proposed method can obtain the DOA estimation without the requirements of spectrum searching or transmit beamspace matrix design, which is different from the conventional DOA estimation techniques. The effectiveness of the proposed method is verified by the simulations.

cs.IT↗

arXiv2020

Twin-Finder: Integrated Reasoning Engine for Pointer-related Code Clone Detection

Detecting code clones is crucial in various software engineering tasks. In particular, code clone detection can have significant uses in the context of analyzing and fixing bugs in large scale applications. However, prior works, such as machine learning-based clone detection, may cause a considerable amount of false positives. In this paper, we propose Twin-Finder, a novel, closed-loop approach for pointer-related code clone detection that integrates machine learning and symbolic execution techniques to achieve precision. Twin-Finder introduces a clone verification mechanism to formally verify if two clone samples are indeed clones and a feedback loop to automatically generated formal rules to tune machine learning algorithm and further reduce the false positives. Our experimental results show that Twin-Finder can swiftly identify up 9X more code clones comparing to a tree-based clone detector, Deckard and remove an average 91.69% false positives.

cs.SE↗

arXiv2020

Algebraic higher symmetry and categorical symmetry -- a holographic and entanglement view of symmetry

We introduce the notion of algebraic higher symmetry, which generalizes higher symmetry and is beyond higher group. We show that an algebraic higher symmetry in a bosonic system in $n$-dimensional space is characterized and classified by a local fusion $n$-category. We find another way to describe algebraic higher symmetry by restricting to symmetric sub Hilbert space where symmetry transformations all become trivial. In this case, algebraic higher symmetry can be fully characterized by a non-invertible gravitational anomaly (i.e. an topological order in one higher dimension). Thus we also refer to non-invertible gravitational anomaly as categorical symmetry to stress its connection to symmetry. This provides a holographic and entanglement view of symmetries. For a system with a categorical symmetry, its gapped state must spontaneously break part (not all) of the symmetry, and the state with the full symmetry must be gapless. Using such a holographic point of view, we obtain (1) the gauging of the algebraic higher symmetry; (2) the classification of anomalies for an algebraic higher symmetry; (3) the equivalence between classes of systems, with different (potentially anomalous) algebraic higher symmetries or different sets of low energy excitations, as long as they have the same categorical symmetry; (4) the classification of gapped liquid phases for bosonic/fermionic systems with a categorical symmetry, as gapped boundaries of a topological order in one higher dimension (that corresponds to the categorical symmetry). This classification includes symmetry protected trivial (SPT) orders and symmetry enriched topological (SET) orders with an algebraic higher symmetry.

cond-mat.str-el↗

arXiv2020

Classification of topological phases with finite internal symmetries in all dimensions

We develop a mathematical theory of symmetry protected trivial (SPT) orders and anomaly-free symmetry enriched topological (SET) orders in all dimensions via two different approaches with an emphasis on the second approach. The first approach is to gauge the symmetry in the same dimension by adding topological excitations as it was done in the 2d case, in which the gauging process is mathematically described by the minimal modular extensions of unitary braided fusion 1-categories. This 2d result immediately generalizes to all dimensions except in 1d, which is treated with special care. The second approach is to use the 1-dimensional higher bulk of the SPT/SET order and the boundary-bulk relation. This approach also leads us to a precise mathematical description and a classification of SPT/SET orders in all dimensions. The equivalence of these two approaches, together with known physical results, provides us with many precise mathematical predictions.

math-ph↗

arXiv2020

Multi-task Learning for Low-resource Second Language Acquisition Modeling

Second language acquisition (SLA) modeling is to predict whether second language learners could correctly answer the questions according to what they have learned. It is a fundamental building block of the personalized learning system and has attracted more and more attention recently. However, as far as we know, almost all existing methods cannot work well in low-resource scenarios due to lacking of training data. Fortunately, there are some latent common patterns among different language-learning tasks, which gives us an opportunity to solve the low-resource SLA modeling problem. Inspired by this idea, in this paper, we propose a novel SLA modeling method, which learns the latent common patterns among different language-learning datasets by multi-task learning and are further applied to improving the prediction performance in low-resource scenarios. Extensive experiments show that the proposed method performs much better than the state-of-the-art baselines in the low-resource scenario. Meanwhile, it also obtains improvement slightly in the non-low-resource scenario.

cs.CL↗

arXiv2020

TNT: Target-driveN Trajectory Prediction

Predicting the future behavior of moving agents is essential for real world applications. It is challenging as the intent of the agent and the corresponding behavior is unknown and intrinsically multimodal. Our key insight is that for prediction within a moderate time horizon, the future modes can be effectively captured by a set of target states. This leads to our target-driven trajectory prediction (TNT) framework. TNT has three stages which are trained end-to-end. It first predicts an agent's potential target states $T$ steps into the future, by encoding its interactions with the environment and the other agents. TNT then generates trajectory state sequences conditioned on targets. A final stage estimates trajectory likelihoods and a final compact set of trajectory predictions is selected. This is in contrast to previous work which models agent intents as latent variables, and relies on test-time sampling to generate diverse trajectories. We benchmark TNT on trajectory prediction of vehicles and pedestrians, where we outperform state-of-the-art on Argoverse Forecasting, INTERACTION, Stanford Drone and an in-house Pedestrian-at-Intersection dataset.

cs.CV↗