Digital Commons Network^™

Online Hyperparameter Optimization For Class-Incremental Learning, Yaoyao Liu, Yingying Li, Bernt Schiele, Qianru Sun

Research Collection School Of Computing and Information Systems

Class-incremental learning (CIL) aims to train a classification model while the number of classes increases phase-by-phase. An inherent challenge of CIL is the stability-plasticity tradeoff, i.e., CIL models should keep stable to retain old knowledge and keep plastic to absorb new knowledge. However, none of the existing CIL models can achieve the optimal tradeoff in different data-receiving settings—where typically the training-from-half (TFH) setting needs more stability, but the training-from-scratch (TFS) needs more plasticity. To this end, we design an online learning method that can adaptively optimize the tradeoff without knowing the setting as a priori. Specifically, we first introduce the …

Web Apis: Features, Issues, And Expectations: A Large-Scale Empirical Study Of Web Apis From Two Publicly Accessible Registries Using Stack Overflow And A User Survey, Neng Zhang, Ying Zou, Xin Xia, David Lo, David Lo, Shanping Li

Research Collection School Of Computing and Information Systems

With the increasing adoption of services-oriented computing and cloud computing technologies, web APIs have become the fundamental building blocks for constructing software applications. Web APIs are developed and published on the internet. The functionality of web APIs can be used to facilitate the development of software applications. There are numerous studies on retrieving and recommending candidate web APIs based on user requirements from a large set of web APIs. However, there are very limited studies on the features of web APIs that make them more likely to be used and the issues of using web APIs in practice. Moreover, users' …

Cross-Domain Graph Anomaly Detection Via Anomaly-Aware Contrastive Alignment, Qizhou Wang, Guansong Pang, Mahsa Salehi, Wray Buntine, Christopher Leckie

Research Collection School Of Computing and Information Systems

Cross-domain graph anomaly detection (CD-GAD) describes the problem of detecting anomalous nodes in an unlabelled target graph using auxiliary, related source graphs with labelled anomalous and normal nodes. Although it presents a promising approach to address the notoriously high false positive issue in anomaly detection, little work has been done in this line of research. There are numerous domain adaptation methods in the literature, but it is difficult to adapt them for GAD due to the unknown distributions of the anomalies and the complex node relations embedded in graph data. To this end, we introduce a novel domain adaptation approach, …

Two Responsibilities To Protect, Patrick Quinton-Brown

Research Collection School of Social Sciences

The purpose of this paper is to re-theorize the evolution of the Responsibility to Protect (RtoP) in the UN through to 2011, the apogee of liberal interventionism in the post-Cold War period. Contrary to a common argument in existing literature, and notwithstanding the adoption of the concept as an annual agenda item of the General Assembly, international contestation is not about implementation as neatly separated from meaning, but rather definition or interpretation. To better understand the boundaries of intergovernmental understanding, we need to interrogate the language or terms of the debate, particularly the ways in which those terms have been …

Bidding Graph Games With Partially-Observable Budgets, Guy Avni, Ismael Jecker, Dorde Zikelic

Research Collection School Of Computing and Information Systems

Two-player zero-sum graph games are a central model, which proceeds as follows. A token is placed on a vertex of a graph, and the two players move it to produce an infinite play, which determines the winner or payoff of the game. Traditionally, the players alternate turns in moving the token. In bidding games, however, the players have budgets and in each turn, an auction (bidding) determines which player moves the token. So far, bidding games have only been studied as fullinformation games. In this work we initiate the study of partial-information bidding games: we study bidding games in which …

Learning Control Policies For Stochastic Systems With Reach-Avoid Guarantees, Dorde Zikelic, Mathias Lechner, A. Thomas Henzinger, Krishnendu Chatterjee

Research Collection School Of Computing and Information Systems

We study the problem of learning controllers for discrete-time non-linear stochastic dynamical systems with formal reach-avoid guarantees. This work presents the first method for providing formal reach-avoid guarantees, which combine and generalize stability and safety guarantees, with a tolerable probability threshold p ∈ [0,1] over the infinite time horizon in general Lipschitz continuous systems. Our method leverages advances in machine learning literature and it represents formal certificates as neural networks. In particular, we learn a certificate in the form of a reach-avoid supermartingale (RASM), a novel notion that we introduce in this work. Our RASMs provide reachability and avoidance guarantees …

Getting Dynamic Implementation To Work, Yi-Chun Chen, Richard Holden, Takashi Kunimoto, Yifei Sun, Tom Wilkening

Research Collection School Of Economics

We develop a new class of two-stage mechanisms, which fully implement any social choice function under initial rationalizability in complete information environments. We show theoretically that our Simultaneous Report (SR) mechanisms are robust to small amounts of incomplete information about the state of nature. We also highlight the robustness of the mechanisms to a wide variety of reasoning processes and behavioral assumptions. We show experimentally that a SR mechanism performs well in inducing truth-telling in both complete and incomplete information environments and that it can induce efficient investment in a two-sided hold-up problem with ex-ante investment. The SR mechanism also …

Do Ambitious Entrepreneurs Benefit More From Training?, Reddi Kotha, Balagopal Vissa, Yimin Lin, Anne-Valérie Corboz

Research Collection Lee Kong Chian School Of Business

Does growth training help entrepreneurs scale-up new ventures? Our field experiment answers this question using data from 181 Singapore-based, early-growth entrepreneurs drawn from a broad range of industry sectors. Treatment content focused on three growth-catalyst tools relevant for formulating and executing innovation-led growth: business-model design, leveraging external networks, building internal teams. Treatment format comprised interactive lecture sessions and workshops on these tools supplemented by personalized coaching in applying the tools to entrepreneurs’ specific challenges. We find that ventures led by entrepreneurs that received training experienced sales growth of 72.5% compared to 30.3% for those in the control condition. Furthermore, ventures …

Marijuana Liberalization And Public Finance: A Capital Market Perspective On The Passage Of Medical Use Laws, Stephanie F. Cheng, Gus De Franco, Pengkai Lin

Research Collection School Of Accountancy

We find that the staggered passage of state-level laws that legalize marijuana for medical use increases states' borrowing costs by 7–9 basis points. Consistent with economic theory on substance use suggesting that marijuana legalization increases local consumption of the drug (by expanding its availability and reducing its perceived risks), we predict and find that increased consumption represents an important mechanism that explains the higher state bond spreads. We also show that following such laws’ passage, states incur higher marijuana-consumption-related expenditures, including for police, corrections, and public welfare.

Executive Function Deficits And Borderline Personality Disorder Symptomatology In A Nonclinical Adult Sample: A Latent Variable Analysis, Keisha D. Veerapandian, Gabriel X. D. Tan, Nadyanna M. Majeed, Andree Hartanto

Research Collection School of Social Sciences

While borderline personality disorder (BPD) symptomatology has been studied extensively in clinical populations, the mechanisms underlying its manifestation in nonclinical populations remain largely understudied. One aspect of BPD symptomatology in nonclinical populations that has not been well studied is cognitive mechanisms, especially in relation to executive functions. To explore the cognitive mechanisms underlying BPD symptomatology in nonclinical populations, we analysed a large-scale dataset of 233 young adults that were administered with nine executive function tasks and BPD symptomatology assessments. Our structural equation modelling did not find any significant relations between latent factors of executive functions and the severity of BPD …

Real-Time Hierarchical Map Segmentation For Coordinating Multi-Robot Exploration, Tianze Luo, Zichen Chen, Budhitama Subagdja, Ah-Hwee Tan

Research Collection School Of Computing and Information Systems

Coordinating a team of autonomous agents to explore an environment can be done by partitioning the map of the environment into segments and allocating the segments as targets for the individual agents to visit. However, given an unknown environment, map segmentation must be conducted in a continuous and incremental manner. In this paper, we propose a novel real-time hierarchical map segmentation method for supporting multi-agent exploration of indoor environments, wherein clusters of regions of segments are formed hierarchically from randomly sampled points in the environment. Each cluster is then assigned with a cost-utility value based on the minimum cost possible …

Deliberate Erring Improves Far Transfer Of Learning More Than Errorless Elaboration And Spotting And Correcting Others' Errors, Sarah Shi Hui Wong

Research Collection School of Social Sciences

Transfer of learning is a fundamental goal of education but is challenging to achieve, especially where far transfer to remote contexts is at stake. How can we improve learners’ flexible application of knowledge to distant domains? In a counterintuitive phenomenon termed the derring effect, deliberately committing and correcting errors in low-stakes contexts enhances learning more than avoiding errors. Whereas this benefit has been demonstrated with tests in domains similar to those in the initial learning task, the present set of three experiments (N = 120) investigated whether deliberate erring boosts far transfer of conceptual knowledge to dissimilar domains. Undergraduates studied …

District-Scale Surface Temperatures Generated From High-Resolution Longitudinal Thermal Infrared Images, Subin Lin, Vasantha Ramani, Miguel Martin, Pandarasamy Arjunan, Adrian Chong, Filip Biljecki, Marcel Ignatius, Clayton Miller, Clayton Miller

Research Collection College of Integrative Studies

This paper describes a dataset collected by infrared thermography, a non-contact, non-intrusive technique to acquire data and analyze the built environment in various aspects. While most studies focus on the city and building scales, an observatory installed on a rooftop provides high temporal and spatial resolution observations with dynamic interactions on the district scale. The rooftop infrared thermography observatory with a multi-modal platform capable of assessing a wide range of dynamic processes in urban systems was deployed in Singapore. It was placed on the top of two buildings that overlook the outdoor context of the National University of Singapore campus. …

Learning Relation Prototype From Unlabeled Texts For Long-Tail Relation Extraction, Yixin Cao, Jun Kuang, Ming Gao, Aoying Zhou, Yonggang Wen, Tat-Seng Chua

Research Collection School Of Computing and Information Systems

Relation Extraction (RE) is a vital step to complete Knowledge Graph (KG) by extracting entity relations from texts. However, it usually suffers from the long-tail issue. The training data mainly concentrates on a few types of relations, leading to the lack of sufficient annotations for the remaining types of relations. In this paper, we propose a general approach to learn relation prototypes from unlabeled texts, to facilitate the long-tail relation extraction by transferring knowledge from the relation types with sufficient training data. We learn relation prototypes as an implicit factor between entities, which reflects the meanings of relations as well …

Covariate Adjustment In Experiments With Matched Pairs, Yuehao Bai, Liang Jiang, Joseph P. Romano, Azeem M. Shaikh, Yichong Zhang

Research Collection School Of Economics

This paper studies inference on the average treatment effect in experiments in which treatment status is determined according to “matched pairs” and it is additionally desired to adjust for observed, baseline covariates to gain further precision. By a “matched pairs” design, we mean that units are sampled i.i.d. from the population of interest, paired according to observed, baseline covariates and finally, within each pair, one unit is selected at random for treatment. Importantly, we presume that not all observed, baseline covariates are used in determining treatment assignment. We study a broad class of estimators based on a “doubly robust” moment …

When Bias Contributes To Variance: True Limit Theory In Functional Coefficient Cointegrating Regression, Peter C. B. Phillips, Ying Wang

Research Collection School Of Economics

Limit distribution theory in the econometric literature for functional coefficient cointegrating regression is incorrect in important ways, influencing rates of convergence, distributional properties, and practical work. The correct limit theory reveals that components from both bias and variance terms contribute to variability in the asymptotics. The errors in the literature arise because random variability in the bias term has been neglected in earlier research. In stationary regression this random variability is of smaller order and can be ignored in asymptotic analysis but not without consequences for finite sample performance. Implications of the findings for rate efficient estimation are discussed. Simulations …

Diagnosing Housing Fever With An Econometric Thermometer, Shuping Shi, Peter C. B. Phillips

Research Collection School Of Economics

Housing fever is a popular term to describe an overheated housing market or housing price bubble. Like other financial asset bubbles, housing fever can inflict harm on the real economy, as indeed the U.S. housing bubble did in the period following 2006 leading up to the general financial crisis and great recession. One contribution that econometricians can make to minimize the harm created by a housing bubble is to provide a quantitative “thermometer” for diagnosing ongoing housing fever. Early diagnosis can enable prompt and effective policy action that reduces long-term damage to the real economy. This paper provides a selective …

Mitigating Popularity Bias For Users And Items With Fairness-Centric Adaptive Recommendation, Zhongzhou Liu, Yuan Fang, Min Wu

Research Collection School Of Computing and Information Systems

Recommendation systems are popular in many domains. Researchers usually focus on the effectiveness of recommendation (e.g., precision) but neglect the popularity bias that may affect the fairness of the recommendation, which is also an important consideration that could influence the benefits of users and item providers. A few studies have been proposed to deal with the popularity bias, but they often face two limitations. Firstly, most studies only consider fairness for one side - either users or items, without achieving fairness jointly for both. Secondly, existing methods are not sufficiently tailored to each individual user or item to cope with …

On Generalized Degree Fairness In Graph Neural Networks, Zemin Liu, Trung Kien Nguyen, Yuan Fang

Research Collection School Of Computing and Information Systems

Conventional graph neural networks (GNNs) are often confronted with fairness issues that may stem from their input, including node attributes and neighbors surrounding a node. While several recent approaches have been proposed to eliminate the bias rooted in sensitive attributes, they ignore the other key input of GNNs, namely the neighbors of a node, which can introduce bias since GNNs hinge on neighborhood structures to generate node representations. In particular, the varying neighborhood structures across nodes, manifesting themselves in drastically different node degrees, give rise to the diverse behaviors of nodes and biased outcomes. In this paper, we first define …

Contrastive Learning Approach To Word-In-Context Task For Low-Resource Languages, Pei-Chi Lo, Yang-Yin Lee, Hsien-Hao Chen, Agus Trisnajaya Kwee, Ee-Peng Lim

Research Collection School Of Computing and Information Systems

Word in context (WiC) task aims to determine whether a target word’s occurrences in two sentences share the same sense. In this paper, we propose a Contrastive Learning WiC (CLWiC) framework to improve the learning of sentence/word representations and classification of target word senses in the sentence pair when performing WiC on lowresource languages. In representation learning, CLWiC trains a pre-trained language model’s ability to cope with lowresource languages using both unsupervised and supervised contrastive learning. The WiC classifier learning further finetunes the language model with WiC classification loss under two classifier architecture options, SGBERT and WiSBERT, which use single-encoder …

Fa3: Fine-Grained Android Application Analysis, Yan Lin, Weng Onn Wong, Debin Gao

Research Collection School Of Computing and Information Systems

Understanding Android applications' behavior is essential to many security applications, e.g., malware analysis. Although many systems have been proposed to perform such dynamic analysis, they are limited by their applicable analysis environment (on device vs. emulator), transparency to subject apps, applicable runtime (Dalvik vs. ART), applicable system stack, or granularity. In this paper, we propose FA3 (Fine-Grained Android Application Analysis), a novel on-device, non-invasive, and fine-grained analysis platform by leveraging existing profiling mechanisms in the Android Runtime (ART) and kernel to inspect method invocations and control-flow transfers for both Java methods and third-party native libraries. FA3 embeds its tracing capability …

Simultaneous Multilateral Search, Sergei Glebkin, Bart Zhou Yueshen, Ji Shen

Research Collection Lee Kong Chian School Of Business

This paper studies simultaneous multilateral search (SMS) in over-the-counter markets: When searching, a customer simultaneously contacts several dealers and trades with the one offering the best quote. Higher search intensity (how often one can search) improves welfare, but higher search capacity (how many dealers one can contact) might be harmful. When the market is in distress, customers might inefficiently favor bilateral bargaining (BB) over SMS. Such a preference for BB speaks to the sluggish adoption of SMS trading, like request-for-quote protocols, in over-the-counter markets. Furthermore, a market-wide shift to SMS may not be socially optimal.

Research@Smu: Sustainable Living, Singapore Management University

Research Collection Office of Research

Sustainable Living is one of the three key priorities of the SMU 2025 Strategy, and the University is committed to develop it into an area of cross-disciplinary strength. The articles in this booklet highlight impactful sustainability research accomplishments at SMU, which spans five broad pillars: Sustainable Business Operations; Sustainable Finance and Impact Assessment; Sustainable Ageing and Wellness; Sustainable Urban Infrastructure; and Sustainable Agro-business and Food Consumption.

Contents:

Sustainable Business Operations

• Managing the Load on Loading Bays
• Going the Last-mile
• Feeding a Growing World
• Pooling the Benefits of Sharing a Ride

Sustainable Finance and Impact Assessment

• When Going Green Becomes a …

Understanding Human-Centred Artificial Intelligence In The Banking Sector, Krishnaraj Arul Obuchettiar, Alan @ Ali Madjelisi Megargel

Research Collection Lee Kong Chian School Of Business

The advent of smart digital devices and social media has shaped how consumers interact and transact with their financial institutions. Consumers increasingly want hyperpersonalised interactions that are more frequent and proactive, while financial institutions have a growing need to cater to consumers’ new demands. Financial institutions, such as banks, continuously adapt to the latest technologies to keep pace with evolving customer behaviours, needs, and experiences. One such emerging technology is artificial intelligence (AI). Many organisations realise the potential of AI; however, a human-centred AI system must be capable of understanding human characteristics and making decisions like humans. This paper presents …

Presidential Economic Approval Rating And The Cross-Section Of Stock Returns, Zilin Chen, Zhi Da, Dashan Huang, Liyao Wang

Research Collection Lee Kong Chian School Of Business

We construct a monthly presidential economic approval rating (PEAR) index from 1981 to 2019, by averaging ratings on the president’s handling of the economy across various national polls. In the cross-section, stocks with high betas to changes in the PEAR index significantly under-perform those with low betas by 1.00% per month in the future, on a risk-adjusted basis. The low PEAR beta premium persists up to one year, and is present in various sub-samples and even in other G7 countries. PEAR beta dynamically reveals a firm’s perceived alignment to the incumbent president’s economic policies and investors seem to misprice such …

Keep Calm And Read Laterally! A Bite-Sized Information Literacy Programme At Singapore Management University Libraries, Zhenyan Li

Research Collection Library

Librarians at Singapore Management University (SMU) Libraries developed an information literacy programme, titled Keep Calm and Read Laterally! as part of the Libraries’ flagship BiteSized Workshop Series. The initiative aimed to further the digital competence of students through advocating critical and effective engagement with digital information. The workshop was highly interactive, incorporating problem-based learning and hands-on group activities, and was split into two parts. The first part helped learners to understand what they read and what types of resources they could use to address various information needs, while the second part guided learners to assess digital information leveraging lateral reading …

The Persistence Of Ethnopopulist Support: The Case Of Rodrigo Duterte's Philippines, Dean C. Dulay, Allen Hicken, Ronald Holmes

Research Collection School of Social Sciences

The past few years have seen an emergence of populist leaders around the world, who have not only accrued but also maintained support despite rampant criticism, governance failures, and the ongoing COVID pandemic. The Philippines' Rodrigo Duterte is the best illustration of this trend, with approval ratings rarely dipping below 80 percent. What explains his high levels of robust public support? We argue that Duterte is an ethnopopulist who uses ethnic appeals in combination with insider vs. outsider rhetoric to garner and maintain public support. Moreover, we argue that ethnic affiliation is a main driver of support for Duterte, and …

Deconstruction Of A Dialogue: Creative Interpretation In Comparative Philosophy, Steven Burik

Research Collection School of Social Sciences

It is common knowledge that Martin Heidegger’s attempts at engaging non-Western philosophy are very much a construct of his own making. This article in no way seeks to disagree with those observations, but argues two things: first, that Heidegger’s “dialogue” with his two main other sources of inspiration, the ancient Greek thinkers and the German poets, is not different in kind or in principle from his engagement with East Asia. One can of course quite easily argue that Heidegger’s main interest was the ancient Greek thinkers, and then the poets, and only lastly Asia. But this hierarchy in preference does …

Can Honesty Reminders Reduce Dishonesty In Budgetary Slack?, Wilfred W. H. Cheng, Chee Yeow Lim, Chi Kwan Yuen

Research Collection School Of Accountancy

This study investigates the effect of honesty reminders on budgetary slack. Based on self-concept maintenance theory, the authors posit that honesty reminders can reduce budgetary slack by making people more aware of their own standards of honesty, resulting in more honest behavior. Using an experimental research design, the authors find evidence that honesty reminders reduce budgetary slack. The authors also find that although penalties can similarly reduce budgetary slack, they tend to cause distrust and resentment from subordinates. Therefore, honesty reminders may be a less costly method than penalties for reducing budgetary slack.

Fortifying The Seams Of Software Systems, Hong Jin Kang

Dissertations and Theses Collection (Open Access)

A seam in software is a place where two components within a software system meet. There are more seams in software now than ever before as modern software systems rely extensively on third-party software components, e.g., libraries. Due to the increasing complexity of software systems, understanding and improving the reliability of these components and their use is crucial. While the use of software components eases the development process, it also introduces challenges due to the interaction between the components.

This dissertation tackles problems associated with software reliability when using third-party software components. Developers write programs that interact with libraries through …