arXiv subjects

Yang Yu

Publications and source records attributed to Yang Yu.

At least 415 records · Page 23Linked to original sources

arXiv2019

Optimal Storage Control for Dynamic Pricing

Renewable energy brings huge uncertainties to the power system, which challenges the traditional power system operation with limited flexible resources. One promising solution is to introduce dynamic pricing to more consumers, which, if designed properly, could enable an active demand side. To further exploit flexibility, in this work, we seek to advice the consumers an optimal online control policy to utilize their storage devices facing dynamic pricing. Towards designing a more adaptive control policy, we devise a data-driven approach to estimating the price distribution. Simulation studies verify the optimality of our proposed schemes.

eess.SY↗

arXiv2019

On Value Discrepancy of Imitation Learning

Imitation learning trains a policy from expert demonstrations. Imitation learning approaches have been designed from various principles, such as behavioral cloning via supervised learning, apprenticeship learning via inverse reinforcement learning, and GAIL via generative adversarial learning. In this paper, we propose a framework to analyze the theoretical property of imitation learning approaches based on discrepancy propagation analysis. Under the infinite-horizon setting, the framework leads to the value discrepancy of behavioral cloning in an order of O((1-γ)^{-2}). We also show that the framework leads to the value discrepancy of GAIL in an order of O((1-γ)^{-1}). It implies that GAIL has less compounding errors than behavioral cloning, which is also verified empirically in this paper. To the best of our knowledge, we are the first one to analyze GAIL's performance theoretically. The above results indicate that the proposed framework is a general tool to analyze imitation learning approaches. We hope our theoretical results can provide insights for future improvements in imitation learning algorithms.

cs.LG↗

arXiv2019

Debugging Crashes using Continuous Contrast Set Mining

Facebook operates a family of services used by over two billion people daily on a huge variety of mobile devices. Many devices are configured to upload crash reports should the app crash for any reason. Engineers monitor and triage millions of crash reports logged each day to check for bugs, regressions, and any other quality problems. Debugging groups of crashes is a manually intensive process that requires deep domain expertise and close inspection of traces and code, often under time constraints. We use contrast set mining, a form of discriminative pattern mining, to learn what distinguishes one group of crashes from another. Prior works focus on discretization to apply contrast mining to continuous data. We propose the first direct application of contrast learning to continuous data, without the need for discretization. We also define a weighted anomaly score that unifies continuous and categorical contrast sets while mitigating bias, as well as uncertainty measures that communicate confidence to developers. We demonstrate the value of our novel statistical improvements by applying it on a challenging dataset from Facebook production logs, where we achieve 40x speedup over baseline approaches using discretization.

cs.SE↗

arXiv2019

An Algorithmic View on Optimal Storage Sizing

Users can arbitrage against Time-of-Use (ToU) pricing with storage by charging in off-peak period and discharge in peak periods. In this paper we design the optimal control policy and the solve optimal investment for general ToU scheme. We formulate the problem as dynamic programming for efficient solution. Our result is feasible facing multi-peaked ToU scheme. Simulation studies examine how the user's cost varies with respect to the user's demand randomness; we also demonstrate the performance of our scheme when aggregating users for extra savings.

eess.SY↗

arXiv2019

Conductor Galloping Prediction on Imbalanced Datasets: SVM with Smart Sampling

Conductor galloping is the high-amplitude, low-frequency oscillation of overhead power lines due to wind. Such movements may lead to severe damages to transmission lines, and hence pose significant risks to the power system operation. In this paper, we target to design a prediction framework for conductor galloping. The difficulty comes from imbalanced dataset as galloping happens rarely. By examining the impacts of data balance and data volume on the prediction performance, we propose to employ proper sample adjustment methods to achieve better performance. Numerical study suggests that using only three features, together with over sampling, the SVM based prediction framework achieves an F_1-score of 98.9%.

eess.SP↗

arXiv2019

Rule Designs for Optimal Online Game Matchmaking

Online games are the most popular form of entertainment among youngsters as well as elders. Recognized as e-Sports, they may become an official part of the Olympic Games by 2020. However, a long waiting time for matchmaking will largely affect players' experiences. We examine different matchmaking mechanisms for 2v2 games. By casting the mechanisms into a queueing theoretic framework, we decompose the rule design process into a sequence of decision making problems, and derive the optimal mechanism with minimum expected waiting time. We further the result by exploring additional static as well as dynamic rule designs' impacts. In the static setting, we consider the game allows players to choose sides before the battle. In the dynamic setting, we consider the game offers multiple zones for players of different skill levels. In both settings, we examine the value of choice-free players. Closed form expressions for the expected waiting time in different settings illuminate the guidelines for online game rule designs.

cs.PF↗

arXiv2019

Signal Combination for Language Identification

Google's multilingual speech recognition system combines low-level acoustic signals with language-specific recognizer signals to better predict the language of an utterance. This paper presents our experience with different signal combination methods to improve overall language identification accuracy. We compare the performance of a lattice-based ensemble model and a deep neural network model to combine signals from recognizers with that of a baseline that only uses low-level acoustic signals. Experimental results show that the deep neural network model outperforms the lattice-based ensemble model, and it reduced the error rate from 5.5% in the baseline to 4.3%, which is a 21.8% relative reduction.

cs.LG↗

arXiv2019

Anisotropies of the g-factor tensor and diamagnetic coefficient in crystal-phase quantum dots in InP nanowires

Crystal-phase low-dimensional structures offer great potential for the implementation of photonic devices of interest for quantum information processing. In this context, unveiling the fundamental parameters of the crystal phase structure is of much relevance for several applications. Here, we report on the anisotropy of the g-factor tensor and diamagnetic coefficient in wurtzite/zincblende (WZ/ZB) crystal-phase quantum dots (QDs) realized in single InP nanowires. The WZ and ZB alternating axial sections in the NWs are identified by high-angle annular dark-field scanning transmission electron microscopy. The electron (hole) g-factor tensor and the exciton diamagnetic coefficients in WZ/ZB crystal-phase QDs are determined through micro-photoluminescence measurements at low temperature (4.2 K) with different magnetic field configurations, and rationalized by invoking the spin-correlated orbital current model. Our work provides key parameters for band gap engineering and spin states control in crystal-phase low-dimensional structures in nanowires.

cond-mat.mes-hall↗

arXiv2019

Hierarchic Neighbors Embedding

Manifold learning now plays a very important role in machine learning and many relevant applications. Although its superior performance in dealing with nonlinear data distribution, data sparsity is always a thorny knot. There are few researches to well handle it in manifold learning. In this paper, we propose Hierarchic Neighbors Embedding (HNE), which enhance local connection by the hierarchic combination of neighbors. After further analyzing topological connection and reconstruction performance, three different versions of HNE are given. The experimental results show that our methods work well on both synthetic data and high-dimensional real-world tasks. HNE develops the outstanding advantages in dealing with general data. Furthermore, comparing with other popular manifold learning methods, the performance on sparse samples and weak-connected manifolds is better for HNE.

cs.LG↗

arXiv2019

Out-of-Domain Detection for Low-Resource Text Classification Tasks

Out-of-domain (OOD) detection for low-resource text classification is a realistic but understudied task. The goal is to detect the OOD cases with limited in-domain (ID) training data, since we observe that training data is often insufficient in machine learning applications. In this work, we propose an OOD-resistant Prototypical Network to tackle this zero-shot OOD detection and few-shot ID classification task. Evaluation on real-world datasets show that the proposed solution outperforms state-of-the-art methods in zero-shot OOD detection task, while maintaining a competitive performance on ID classification task.

cs.CL↗

arXiv2019

LabelECG: A Web-based Tool for Distributed Electrocardiogram Annotation

Electrocardiography plays an essential role in diagnosing and screening cardiovascular diseases in daily healthcare. Deep neural networks have shown the potentials to improve the accuracies of arrhythmia detection based on electrocardiograms (ECGs). However, more ECG records with ground truth are needed to promote the development and progression of deep learning techniques in automatic ECG analysis. Here we propose a web-based tool for ECG viewing and annotating, LabelECG. With the facilitation of unified data management, LabelECG is able to distribute large cohorts of ECGs to dozens of technicians and physicians, who can simultaneously make annotations through web-browsers on PCs, tablets and cell phones. Along with the doctors from four hospitals in China, we applied LabelECG to support the annotations of about 15,000 12-lead resting ECG records in three months. These annotated ECGs have successfully supported the First China ECG intelligent Competition. La-belECG will be freely accessible on the Internet to support similar researches, and will also be upgraded through future works.

cs.DB↗

arXiv2019

Note on "Experimental Measurement of Quantum Metric Tensor and Related Topological Phase Transition with a Superconducting Qubit"

In the paper [X. Tan et al., Phys. Rev. Lett. 122, 210401 (2019)], we have studied the Euler characteristic number $χ$ for a two-band model; However, the calculated $χ$ there is not an integer when the parameter $|h|>1$ and then may not be considered as a suitable topological number for the system. In this note, we find that the Bloch vectors of the ground state for $|h|>1$ do not cover the whole Bloch sphere and thus the Euler characteristic number should be effectively associated with the manifold of a disk, rather than a sphere. After taking into account the boundary contribution, we derive the correct Euler characteristic number. Unfortunately, the Euler characteristic number does not change when crossing the critical points $|h|=1$ and thus can not be used to characterize the topological phase transition of the present model.

quant-ph↗

arXiv2019

Topologically Protected Photonic Modes in Composite Quantum Hall/Quantum Spin Hall Waveguides

Photonic topological systems, the electromagnetic analog of the topological materials in condensed matter physics, create many opportunities to design optical devices with novel properties. We present an experimental realization of the bi-anisotropic meta waveguide photonic system replicating both quantum Hall (QH) and quantum spin-Hall (QSH) topological insulating phases. With careful design, a composite QH-QSH photonic topological material is created and experimentally shown to support reflection-free edgemodes, a heterogeneous topological structure that is unprecedented in condensed matter physics. The effective spin degree of freedom of such topologically protected modes determines their unique pathways through these systems, free from backscattering and able to travel around sharp corners. {As an example of their novel properties, we experimentally demonstrate reflection-less photonic devices including a 2-port isolator, a unique 3-port topological device, and a full 4-port circulator based on composite QH and QSH structures

cond-mat.mes-hall↗

arXiv2019

Key Ingredients of Self-Driving Cars

Over the past decade, many research articles have been published in the area of autonomous driving. However, most of them focus only on a specific technological area, such as visual environment perception, vehicle control, etc. Furthermore, due to fast advances in the self-driving car technology, such articles become obsolete very fast. In this paper, we give a brief but comprehensive overview on key ingredients of autonomous cars (ACs), including driving automation levels, AC sensors, AC software, open source datasets, industry leaders, AC applications and existing challenges.

cs.RO↗

arXiv2019

Towards AutoML in the presence of Drift: first results

Research progress in AutoML has lead to state of the art solutions that can cope quite wellwith supervised learning task, e.g., classification with AutoSklearn. However, so far thesesystems do not take into account the changing nature of evolving data over time (i.e., theystill assume i.i.d. data); even when this sort of domains are increasingly available in realapplications (e.g., spam filtering, user preferences, etc.). We describe a first attempt to de-velop an AutoML solution for scenarios in which data distribution changes relatively slowlyover time and in which the problem is approached in a lifelong learning setting. We extendAuto-Sklearn with sound and intuitive mechanisms that allow it to cope with this sort ofproblems. The extended Auto-Sklearn is combined with concept drift detection techniquesthat allow it to automatically determine when the initial models have to be adapted. Wereport experimental results in benchmark data from AutoML competitions that adhere tothis scenario. Results demonstrate the effectiveness of the proposed methodology.

cs.LG↗

arXiv2019

Transition radiation in photonic topological crystals: quasi-resonant excitation of robust edge states by a moving charge

We demonstrate, theoretically and experimentally, that a traveling electric charge passing from one photonic crystal into another generates edge waves -- electromagnetic modes with frequencies inside the common photonic bandgap localized at the interface -- via a process of transition edge-wave radiation (TER). A simple and intuitive expression for the TER spectral density is derived and then applied to a specific structure: two interfacing photonic topological insulators with opposite spin-Chern indices. We show that TER breaks the time-reversal symmetry and enables valley- and spin-polarized generation of topologically protected edge waves propagating in one or both directions along the interface. Experimental measurements at the Argonne Wakefield Accelerator Facility are consistent with the excitation and localization of the edge waves. The concept of TER paves the way for novel particle accelerators and detectors.

physics.optics↗

arXiv2019

Environment Reconstruction with Hidden Confounders for Reinforcement Learning based Recommendation

Reinforcement learning aims at searching the best policy model for decision making, and has been shown powerful for sequential recommendations. The training of the policy by reinforcement learning, however, is placed in an environment. In many real-world applications, however, the policy training in the real environment can cause an unbearable cost, due to the exploration in the environment. Environment reconstruction from the past data is thus an appealing way to release the power of reinforcement learning in these applications. The reconstruction of the environment is, basically, to extract the casual effect model from the data. However, real-world applications are often too complex to offer fully observable environment information. Therefore, quite possibly there are unobserved confounding variables lying behind the data. The hidden confounder can obstruct an effective reconstruction of the environment. In this paper, by treating the hidden confounder as a hidden policy, we propose a deconfounded multi-agent environment reconstruction (DEMER) approach in order to learn the environment together with the hidden confounder. DEMER adopts a multi-agent generative adversarial imitation learning framework. It proposes to introduce the confounder embedded policy, and use the compatible discriminator for training the policies. We then apply DEMER in an application of driver program recommendation. We firstly use an artificial driver program recommendation environment, abstracted from the real application, to verify and analyze the effectiveness of DEMER. We then test DEMER in the real application of Didi Chuxing. Experiment results show that DEMER can effectively reconstruct the hidden confounder, and thus can build the environment better. DEMER also derives a recommendation policy with a significantly improved performance in the test phase of the real application.

cs.LG↗

arXiv2019

Experimental Measurement of the Quantum Metric Tensor and Related Topological Phase Transition with a Superconducting Qubit

Berry curvature is an imaginary component of the quantum geometric tensor (QGT) and is well studied in many branches of modern physics; however, the quantum metric as a real component of the QGT is less explored. Here, by using tunable superconducting circuits, we experimentally demonstrate two methods to directly measure the quantum metric tensor for characterizing the geometry and topology of underlying quantum states in parameter space. The first method is to probe the transition probability after a sudden quench, and the second one is to detect the excitation rate under weak periodic driving. Furthermore, based on quantum-metric and Berry-curvature measurements, we explore a topological phase transition in a simulated time-reversal-symmetric system, which is characterized by the Euler characteristic number instead of the Chern number. The work opens up a unique approach to explore the topology of quantum states with the QGT.

quant-ph↗