Open Access. Powered by Scholars. Published by Universities.®

Digital Commons Network™

Open Access. Powered by Scholars. Published by Universities.®

Singapore Management University

Discipline
Keyword
Publication Year
Publication
Publication Type

Articles 2821 - 2850 of 8706

Full-Text Articles in Entire DC Network

Quasi-Bayesian Inference For Production Frontiers, Xiaobin Liu, Thomas Tao Yang, Yichong Zhang Dec 2020

Quasi-Bayesian Inference For Production Frontiers, Xiaobin Liu, Thomas Tao Yang, Yichong Zhang

Research Collection School Of Economics

We propose a quasi-Bayesian method to conduct inference for the production frontier. This approach combines multiple first-stage extreme quantile estimates by the quasi-Bayesian method to produce the point estimate and confidence interval for the production frontier. We show the asymptotic properties of the proposed estimator and the validity of the inference procedure. The finite sample performance of our method is illustrated through simulations and an empirical application.

Go to article

Density Of Demand And The Consumer Benefit From Uber, Matthew H. Shapiro Dec 2020

Density Of Demand And The Consumer Benefit From Uber, Matthew H. Shapiro

Research Collection School Of Economics

Uber has attracted the attention of economists and policy makers for its innovations in the taxicab market and its potential for significant consumer welfare gains. The size of this gain depends in part on whether these innovations permit transactions previously costly or infeasible. Using New York City — the largest taxi market in the country — as its context, this paper estimates the level of any technological advantage Uber has over hail taxis in matching to consumers. I combine publicly available transportation data with data scraped from Uber and traffic cameras to estimate a model of the demand for transportation …

Go to article

Secure Answer Book And Automatic Grading, Manoj Thulasidas Dec 2020

Secure Answer Book And Automatic Grading, Manoj Thulasidas

Research Collection School Of Computing and Information Systems

In response to the growing need to perform assessments online, we have developed a secure answer book, as well as a tool for automatically grading it for our course on spread- sheet modeling. We applied these techniques to a cohort of about 160 students who took the course last term. In this paper, we describe the design, implementation and the techniques employed to enhance both the security of the answer book and the ease, accuracy and consistency of grading. In addition, we summarize the experience and takeaways, both from the instructor and the student perspectives. Although the answer book and …

Go to article

Jointly Optimizing Sensing Pipelines For Multimodal Mixed Reality Interaction, Darshana Rathnayake, Ashen De Silva, Dasun Puwakdandawa, Lakmal Meegahapola, Archan Misra, Indika Perera Dec 2020

Jointly Optimizing Sensing Pipelines For Multimodal Mixed Reality Interaction, Darshana Rathnayake, Ashen De Silva, Dasun Puwakdandawa, Lakmal Meegahapola, Archan Misra, Indika Perera

Research Collection School Of Computing and Information Systems

Natural human interactions for Mixed Reality Applications are overwhelmingly multimodal: humans communicate intent and instructions via a combination of visual, aural and gestural cues. However, supporting low-latency and accurate comprehension of such multimodal instructions (MMI), on resource-constrained wearable devices, remains an open challenge, especially as the state-of-the-art comprehension techniques for each individual modality increasingly utilize complex Deep Neural Network models. We demonstrate the possibility of overcoming the core limitation of latency--vs.--accuracy tradeoff by exploiting cross-modal dependencies -- i.e., by compensating for the inferior performance of one model with an increased accuracy of more complex model of a different modality. We …

Go to article

Does The Fossil Fuel Divestment Movement Impact New Oil And Gas Fundraising?, Theodor Florian Cojoianu, Francisco Ascui, Gordon Clark, Andreas Hoepnner, Dariusz Wójcik Dec 2020

Does The Fossil Fuel Divestment Movement Impact New Oil And Gas Fundraising?, Theodor Florian Cojoianu, Francisco Ascui, Gordon Clark, Andreas Hoepnner, Dariusz Wójcik

Research Collection College of Integrative Studies

This article explores whether increasing fossil fuel divestment commitments are related to the reduction of capital flows into the oil and gas sector, based on an analysis of syndicated lending, equity and bond underwriting across 33 countries from 2000 to 2015. We find that increasing oil and gas divestment pledges in a country are associated with lower capital flows to domestic oil and gas companies. This effect is enhanced in more stringent environmental policy regimes and diminished in countries which heavily subsidise fossil fuels. However, the divestment movement may have an unintended effect, insofar as domestic banks situated in countries …

Go to article

The Future Of Reorganization Procedures In The Era Of Pre-Insolvency Law, Aurelio Gurrea-Martinez Dec 2020

The Future Of Reorganization Procedures In The Era Of Pre-Insolvency Law, Aurelio Gurrea-Martinez

Research Collection Yong Pung How School Of Law

Several countries and regions around the world, including Singapore, the United Kingdom, and the European Union, are amending their restructuring framework to implement a pre-insolvency mechanism that includes most of the features that exist in the US Chapter 11 reorganization procedure. However, unlike what happens in the United States, where unsuccessful reorganizations lead to Chapter 7 liquidations, companies using this ‘de facto Chapter 11’ (DFCH11) are still allowed to use formal reorganization procedures. This article argues that, while the rise of the DFCH11 is not necessarily undesirable provided that various protections are put in place, jurisdictions implementing this restructuring tool …

Go to article

Form And Substance In Singapore Constitutional And Administrative Law, Kenny Chng Dec 2020

Form And Substance In Singapore Constitutional And Administrative Law, Kenny Chng

Research Collection Yong Pung How School Of Law

This paper proposes to study constitutional and administrative law in Singapore through the lenses of Atiyah’s and Summers’ concepts of form and substance in order to discern fruitful avenues for the development of Singapore constitutional and administrative law. While the concepts of form and substance in the context of constitutional law are often associated with constitutional interpretation, they can also be fruitfully applied to other areas of constitutional and administrative law to shed light on the potential trajectories of Singapore law. The intent of this paper is to apply Atiyah’s and Summers’ concepts of form and substance to Singapore constitutional …

Go to article

The Future Is Urban: The Progressive Renaissance Of The City In Eu Law, De Maartje Visser Dec 2020

The Future Is Urban: The Progressive Renaissance Of The City In Eu Law, De Maartje Visser

Research Collection Yong Pung How School Of Law

For much of the European integration process, local authorities have been on the legal margins. Yet many amongst this group, and cities in particular, consider themselves as important players in realising the Union’s overarching policy objectives. This view is slowly but surely fi nding traction with the EU’s political institutions. This article suggests that the future architecture of the European Union’s (EU’s) operating system will evince a rapprochement between the socio-economic clout of local authorities, notably cities, and their legal-political recognition at Union level. It further suggests that there is room for greater conceptual clarity along two lines when interrogating …

Go to article

Improving Gan Training With Probability Ratio Clipping And Sample Reweighting, Yue Wu, Pan Zhou, Andrew Wilson Gordon, Eric Xing, Zhiting Hu Dec 2020

Improving Gan Training With Probability Ratio Clipping And Sample Reweighting, Yue Wu, Pan Zhou, Andrew Wilson Gordon, Eric Xing, Zhiting Hu

Research Collection School Of Computing and Information Systems

Despite success on a wide range of problems related to vision, generative adversarial networks (GANs) often suffer from inferior performance due to unstable training, especially for text generation. To solve this issue, we propose a new variational GAN training framework which enjoys superior training stability. Our approach is inspired by a connection of GANs and reinforcement learning under a variational perspective. The connection leads to (1) probability ratio clipping that regularizes generator training to prevent excessively large updates, and (2) a sample re-weighting mechanism that improves discriminator training by downplaying bad-quality fake samples. Moreover, our variational GAN framework can provably …

Go to article

Audee: Automated Testing For Deep Learning Frameworks, Qianyu Guo, Xiaofei Xie, Yi Li, Xiaoyu Zhang, Yang Liu, Xiaohong Li, Chao Shen Dec 2020

Audee: Automated Testing For Deep Learning Frameworks, Qianyu Guo, Xiaofei Xie, Yi Li, Xiaoyu Zhang, Yang Liu, Xiaohong Li, Chao Shen

Research Collection School Of Computing and Information Systems

Deep learning (DL) has been applied widely, and the quality of DL system becomes crucial, especially for safety-critical applications. Existing work mainly focuses on the quality analysis of DL models, but lacks attention to the underlying frameworks on which all DL models depend. In this work, we propose Audee, a novel approach for testing DL frameworks and localizing bugs. Audee adopts a search-based approach and implements three different mutation strategies to generate diverse test cases by exploring combinations of model structures, parameters, weights and inputs. Audee is able to detect three types of bugs: logical bugs, crashes and Not-a-Number (NaN) …

Go to article

Enabling Collaborative Video Sensing At The Edge Through Convolutional Sharing, Kasthuri Jayarajah, Wanniarachchige Dhanuja Tharith Wanniarachchi, Archan Misra Dec 2020

Enabling Collaborative Video Sensing At The Edge Through Convolutional Sharing, Kasthuri Jayarajah, Wanniarachchige Dhanuja Tharith Wanniarachchi, Archan Misra

Research Collection School Of Computing and Information Systems

While Deep Neural Network (DNN) models have provided remarkable advances in machine vision capabilities, their high computational complexity and model sizes present a formidable roadblock to deployment in AIoT-based sensing applications. In this paper, we propose a novel paradigm by which peer nodes in a network can collaborate to improve their accuracy on person detection, an exemplar machine vision task. The proposed methodology requires no re-training of the DNNs and incurs minimal processing latency as it extracts scene summaries from the collaborators and injects back into DNNs of the reference cameras, on-the-fly. Early results show promise with improvements in recall …

Go to article

Interventional Few-Shot Learning, Zhongqi Yue, Zhang Hanwang, Qianru Sun, Xian-Sheng Hua Dec 2020

Interventional Few-Shot Learning, Zhongqi Yue, Zhang Hanwang, Qianru Sun, Xian-Sheng Hua

Research Collection School Of Computing and Information Systems

We uncover an ever-overlooked deficiency in the prevailing Few-Shot Learning (FSL) methods: the pre-trained knowledge is indeed a confounder that limits the performance. This finding is rooted from our causal assumption: a Structural Causal Model (SCM) for the causalities among the pre-trained knowledge, sample features, and labels. Thanks to it, we propose a novel FSL paradigm: Interventional Few-Shot Learning (IFSL). Specifically, we develop three effective IFSL algorithmic implementations based on the backdoor adjustment, which is essentially a causal intervention towards the SCM of many-shot learning: the upper-bound of FSL in a causal view. It is worth noting that the contribution …

Go to article

A Bert-Based Dual Embedding Model For Chinese Idiom Prediction, Minghuan Tan, Jing Jiang Dec 2020

A Bert-Based Dual Embedding Model For Chinese Idiom Prediction, Minghuan Tan, Jing Jiang

Research Collection School Of Computing and Information Systems

Chinese idioms are special fixed phrases usually derived from ancient stories, whose meanings are oftentimes highly idiomatic and non-compositional. The Chinese idiom prediction task is to select the correct idiom from a set of candidate idioms given a context with a blank. We propose a BERT-based dual embedding model to encode the contextual words as well as to learn dual embeddings of the idioms. Specifically, we first match the embedding of each candidate idiom with the hidden representation corresponding to the blank in the context. We then match the embedding of each candidate idiom with the hidden representations of all …

Go to article

A Deep Learning Framework Supporting Model Ownership Protection And Traitor Tracing, Guowen Xu, Hongwei Li, Yuan Zhang, Xiaodong Lin, Robert H. Deng, Xuemin (Sherman) Shen Dec 2020

A Deep Learning Framework Supporting Model Ownership Protection And Traitor Tracing, Guowen Xu, Hongwei Li, Yuan Zhang, Xiaodong Lin, Robert H. Deng, Xuemin (Sherman) Shen

Research Collection School Of Computing and Information Systems

Cloud-based deep learning (DL) solutions have been widely used in applications ranging from image recognition to speech recognition. Meanwhile, as commercial software and services, such solutions have raised the need for intellectual property rights protection of the underlying DL models. Watermarking is the mainstream of existing solutions to address this concern, by primarily embedding pre-defined secrets in a model's training process. However, existing efforts almost exclusively focus on detecting whether a target model is pirated, without considering traitor tracing. In this paper, we present SecureMark_DL, which enables a model owner to embed a unique fingerprint for every customer within parameters …

Go to article

Covid-19, Coronavirus, Wuhan Virus, Or China Virus? Understanding How To “Do No Harm" When Naming An Infectious Disease, Theodore C. Masters-Waage, Nilotpal Jha, Jochen Reb Dec 2020

Covid-19, Coronavirus, Wuhan Virus, Or China Virus? Understanding How To “Do No Harm" When Naming An Infectious Disease, Theodore C. Masters-Waage, Nilotpal Jha, Jochen Reb

Research Collection Lee Kong Chian School Of Business

When labeling an infectious disease, officially sanctioned scientific names, e.g., “H1N1 virus,” are recommended over place-specific names, e.g., “Spanish flu.” This is due to concerns from policymakers and the WHO that the latter might lead to unintended stigmatization. However, with little empirical support for such negative consequences, authorities might be focusing on limited resources on an overstated issue. This paper empirically investigates the impact of naming against the current backdrop of the 2019-2020 pandemic.

Go to article

Data Driven Value-At-Risk Forecasting Using A Svr-Garch-Kde Hybrid, Marius Lux, Wolfgang Karl Hardle, Stefan Lessmann Nov 2020

Data Driven Value-At-Risk Forecasting Using A Svr-Garch-Kde Hybrid, Marius Lux, Wolfgang Karl Hardle, Stefan Lessmann

Sim Kee Boon Institute for Financial Economics

Appropriate risk management is crucial to ensure the competitiveness of financial institutions and the stability of the economy. One widely used financial risk measure is value-at-risk (VaR). VaR estimates based on linear and parametric models can lead to biased results or even underestimation of risk due to time varying volatility, skewness and leptokurtosis of financial return series. The paper proposes a nonlinear and nonparametric framework to forecast VaR that is motivated by overcoming the disadvantages of parametric models with a purely data driven approach. Mean and volatility are modeled via support vector regression (SVR) where the volatility model is motivated …

Go to article

Macroeconomic Stabilization In The Digital Age, John Beirne, David Fernandez Nov 2020

Macroeconomic Stabilization In The Digital Age, John Beirne, David Fernandez

Research Collection Lee Kong Chian School Of Business

Macroeconomic Stabilization in the Digital Age provides insights into factors affecting the macroeconomic management of the economy in the digital age. Policy makers need to be aware of the increasing prominence of the digital economy and digital finance and seek to better understand how continued digitalization will affect policies aimed at managing the economy. For emerging market economies (EMEs), macroeconomic policy challenges have been exacerbated by the digital finance revolution in the aftermath of the global financial crisis and the coronavirus disease (COVID-19) pandemic, when many EMEs experienced large and volatile capital flows. Policy makers must also navigate through fluctuating …

Go to article

When And Why Narcissists Exhibit Greater Hindsight Bias And Less Perceived Learning, Satoris S. Howes, Edgar E. Kausel, Alexander T. Jackson, Jochen Reb Nov 2020

When And Why Narcissists Exhibit Greater Hindsight Bias And Less Perceived Learning, Satoris S. Howes, Edgar E. Kausel, Alexander T. Jackson, Jochen Reb

Research Collection Lee Kong Chian School Of Business

The present research sought to examine the impact of narcissism, prediction accuracy, and should counterfactual thinking—which includes thoughts such as “I should have done something different”—on hindsight bias (the tendency to exaggerate in hindsight what one knew in foresight) and perceived learning. To test these effects, we conducted four studies (total n = 727). First, in Study 1 we examined a moderated mediation model, in which should counterfactual thinking mediates the relation between narcissism and hindsight bias, and this mediation is moderated by prediction accuracy such that the relationship is negative when predictions are accurate and positive when predictions are …

Go to article

Cache-Based Optimization For Block Commit Of Hyperledger Fabric, Qinqi Xu, Yimin Lin, Qingshan Jiang, Mengqiu Zhang Nov 2020

Cache-Based Optimization For Block Commit Of Hyperledger Fabric, Qinqi Xu, Yimin Lin, Qingshan Jiang, Mengqiu Zhang

Research Collection Lee Kong Chian School Of Business

Blockchain technology has been applied to various domains in recent years, and Hyperledger Fabric is one of the most popular blockchain systems. This paper conducts a fine analysis on the block commit of Hyperledger Fabric through latency evaluation as well as flame graph. To overcome the bottlenecks pointed out by the analysts, this paper proposes two cache-based schemata. In particular, we add a cache for index and state storage to accelerate commit. We also cache deserialized block globally for block delivery to avoid repeat operations. The experiments reveal performance improvement of 2×. Notably, our optimizations can also be implemented easily …

Go to article

Tangi: Tangible Proxies For Embodied Object Exploration And Manipulation In Virtual Reality, Martin Feick, Scott Bateman, Anthony Tang, Anthony Tang Nov 2020

Tangi: Tangible Proxies For Embodied Object Exploration And Manipulation In Virtual Reality, Martin Feick, Scott Bateman, Anthony Tang, Anthony Tang

Research Collection School Of Computing and Information Systems

Exploring and manipulating complex virtual objects is challenging due to limitations of conventional controllers and free-hand interaction techniques. We present the TanGi toolkit which enables novices to rapidly build physical proxy objects using Composable Shape Primitives. TanGi also provides Manipulators allowing users to build objects including movable parts, making them suitable for rich object exploration and manipulation in VR. With a set of different use cases and applications we show the capabilities of the TanGi toolkit and evaluate its use. In a study with 16 participants, we demonstrate that novices can quickly build physical proxy objects using the Composable Shape …

Go to article

Cost-Sensitive Deep Forest For Price Prediction, Chao Ma, Zhenbing Liu, Zhiguang Cao, Wen Song, Jie Zhang, Weiliang Zeng Nov 2020

Cost-Sensitive Deep Forest For Price Prediction, Chao Ma, Zhenbing Liu, Zhiguang Cao, Wen Song, Jie Zhang, Weiliang Zeng

Research Collection School Of Computing and Information Systems

For many real-world applications, predicting a price range is more practical and desirable than predicting a concrete value. In this case, price prediction can be regarded as a classification problem. Although deep forest is recognized as the best solution to many classification problems, a crucial issue limits its direct application to price prediction, i.e., it treated all the misclassifications equally no matter how far away they are from the real classes, since their impacts on the accuracy are the same. This is unreasonable to price prediction as the misclassification should be as close to the real price range as possible …

Go to article

Coupled Hierarchical Transformer For Stance-Aware Rumor Verification In Social Media Conversations, Jianfei Yu, Jing Jiang, Ling Min Serena Khoo, Hai Leong Chieu, Rui Xia Nov 2020

Coupled Hierarchical Transformer For Stance-Aware Rumor Verification In Social Media Conversations, Jianfei Yu, Jing Jiang, Ling Min Serena Khoo, Hai Leong Chieu, Rui Xia

Research Collection School Of Computing and Information Systems

The prevalent use of social media enables rapid spread of rumors on a massive scale, which leads to the emerging need of automatic rumor verification (RV). A number of previous studies focus on leveraging stance classification to enhance RV with multi-task learning (MTL) methods. However, most of these methods failed to employ pre-trained contextualized embeddings such as BERT, and did not exploit inter-task dependencies by using predicted stance labels to improve the RV task. Therefore, in this paper, to extend BERT to obtain thread representations, we first propose a Hierarchical Transformer1 , which divides each long thread into shorter subthreads, …

Go to article

Deep-Learning-Based App Sensitive Behavior Surveillance For Android Powered Cyber-Physical Systems, Haoyu Ma, Jianwen Tian, Kefan Qiu, David Lo, Debin Gao, Daoyuan Wu, Chunfu Jia, Thar Baker Nov 2020

Deep-Learning-Based App Sensitive Behavior Surveillance For Android Powered Cyber-Physical Systems, Haoyu Ma, Jianwen Tian, Kefan Qiu, David Lo, Debin Gao, Daoyuan Wu, Chunfu Jia, Thar Baker

Research Collection School Of Computing and Information Systems

Android as an operating system is now increasingly being adopted in industrial information systems, especially with Cyber-Physical Systems (CPS). This also puts Android devices onto the front line of handling security-related data and conducting sensitive behaviors, which could be misused by the increasing number of polymorphic and metamorphic malicous applications targeting the platform. The existence of such malware threats therefore call for more accurate identification and surveillance of sensitive Android app behaviors, which is essential to the security of CPS and IoT devices powered by Android. Nevertheless, achieving dynamic app behavior monitoring and identification on real CPS powered by Android …

Go to article

Cross-Thought For Sentence Encoder Pre-Training, Shuohang Wang, Yuwei Fang, Siqi Sun, Zhe Gan, Yu Cheng, Jingjing Liu, Jing Jiang Nov 2020

Cross-Thought For Sentence Encoder Pre-Training, Shuohang Wang, Yuwei Fang, Siqi Sun, Zhe Gan, Yu Cheng, Jingjing Liu, Jing Jiang

Research Collection School Of Computing and Information Systems

In this paper, we propose Cross-Thought, a novel approach to pre-training sequence encoder, which is instrumental in building reusable sequence embeddings for large-scale NLP tasks such as question answering. Instead of using the original signals of full sentences, we train a Transformer-based sequence encoder over a large set of short sequences, which allows the model to automatically select the most useful information for predicting masked words. Experiments on question answering and textual entailment tasks demonstrate that our pre-trained encoder can outperform state-of-the-art encoders trained with continuous sentence signals as well as traditional masked language modeling baselines. Our proposed approach also …

Go to article

Multi-User Verifiable Searchable Symmetric Encryption For Cloud Storage, Xueqiao Liu, Guomin Yang, Guomin Yang Nov 2020

Multi-User Verifiable Searchable Symmetric Encryption For Cloud Storage, Xueqiao Liu, Guomin Yang, Guomin Yang

Research Collection School Of Computing and Information Systems

In a cloud data storage system, symmetric key encryption is usually used to encrypt files due to its high efficiency. In order allow the untrusted/semi-trusted cloud storage server to perform searching over encrypted data while maintaining data confidentiality, searchable symmetric encryption (SSE) has been proposed. In a typical SSE scheme, a users stores encrypted files on a cloud storage server and later can retrieve the encrypted files containing specific keywords. The basic security requirement of SSE is that the cloud server learns no information about the files or the keywords during the searching process. Some SSE schemes also offer additional …

Go to article

A Matter Of Life In Debt, Deepika Deshpande Nov 2020

A Matter Of Life In Debt, Deepika Deshpande

Asian Management Insights

Moving away from using debt to tackle economic downturns.

Go to article

Attribute-Based Keyword Search Over Hierarchical Data In Cloud Computing, Yinbin Miao, Jianfeng Ma, Ximeng Liu, Xinghua Li, Qi Jiang, Junwei Zhang Nov 2020

Attribute-Based Keyword Search Over Hierarchical Data In Cloud Computing, Yinbin Miao, Jianfeng Ma, Ximeng Liu, Xinghua Li, Qi Jiang, Junwei Zhang

Research Collection School Of Computing and Information Systems

Searchable encryption (SE) has been a promising technology which allows users to perform search queries over encrypted data. However, the most of existing SE schemes cannot deal with the shared records that have hierarchical structures. In this paper, we devise a basic cryptographic primitive called as attribute-based keyword search over hierarchical data (ABKS-HD) scheme by using the ciphertext-policy attribute-based encryption (CP-ABE) technique, but this basic scheme cannot satisfy all the desirable requirements of cloud systems. The facts that the single keyword search will yield many irrelevant search results and the revoked users can access the unauthorized data with the old …

Go to article

Erica: Enabling Real-Time Mistake Detection And Corrective Feedback For Free-Weights Exercises, Meeralakshmi Radhakrishnan, Darshana Rathnayake, Koon Han Ong, Inseok Hwang, Archan Misra Nov 2020

Erica: Enabling Real-Time Mistake Detection And Corrective Feedback For Free-Weights Exercises, Meeralakshmi Radhakrishnan, Darshana Rathnayake, Koon Han Ong, Inseok Hwang, Archan Misra

Research Collection School Of Computing and Information Systems

We present ERICA, a digital personal trainer for users performing free weights exercises, with two key differentiators: (a) First, unlike prior approaches that either require multiple on-body wearables or specialized infrastructural sensing, ERICA uses a single in-ear "earable" device (piggybacking on a form factor routinely used by millions of gym-goers) and a simple inertial sensor mounted on each weight equipment; (b) Second, unlike prior work that focuses primarily on quantifying a workout, ERICA additionally identifies a variety of fine-grained exercising mistakes and delivers real-time, in-situ corrective instructions. To achieve this, we (a) design a robust approach for user-equipment association that …

Go to article

Reducing Estimation Bias Via Triplet-Average Deep Deterministic Policy Gradient, Dongming Wu, Xingping Dong, Jianbing Shen, Steven C. H. Hoi Nov 2020

Reducing Estimation Bias Via Triplet-Average Deep Deterministic Policy Gradient, Dongming Wu, Xingping Dong, Jianbing Shen, Steven C. H. Hoi

Research Collection School Of Computing and Information Systems

The overestimation caused by function approximation is a well-known property in Q-learning algorithms, especially in single-critic models, which leads to poor performance in practical tasks. However, the opposite property, underestimation, which often occurs in Q-learning methods with double critics, has been largely left untouched. In this article, we investigate the underestimation phenomenon in the recent twin delay deep deterministic actor-critic algorithm and theoretically demonstrate its existence. We also observe that this underestimation bias does indeed hurt performance in various experiments. Considering the opposite properties of single-critic and double-critic methods, we propose a novel triplet-average deep deterministic policy gradient algorithm that …

Go to article

Boosting Privately: Federated Extreme Gradient Boosting For Mobile Crowdsensing, Yang Liu, Zhuo Ma, Ximeng Liu, Siqi Ma, Surya Nepal, Robert H. Deng, Kui Ren Nov 2020

Boosting Privately: Federated Extreme Gradient Boosting For Mobile Crowdsensing, Yang Liu, Zhuo Ma, Ximeng Liu, Siqi Ma, Surya Nepal, Robert H. Deng, Kui Ren

Research Collection School Of Computing and Information Systems

Recently, Google and other 24 institutions proposed a series of open challenges towards federated learning (FL), which include application expansion and homomorphic encryption (HE). The former aims to expand the applicable machine learning models of FL. The latter focuses on who holds the secret key when applying HE to FL. For the naive HE scheme, the server is set to master the secret key. Such a setting causes a serious problem that if the server does not conduct aggregation before decryption, a chance is left for the server to access the user’s update. Inspired by the two challenges, we propose …

Go to article

First Prev Next Last