Search

Linear scalarization, i.e., combining all loss functions by a weighted sum, has been the default choice in the literature of multi-task learning (MTL) since its inception. In recent years, there is a surge of interest in developing Specialized Multi-Task Optimizers (SMTOs) that treat MTL as a multi-objective optimization problem. However, it remains open whether there is a fundamental advantage of SMTOs over scalarization. In fact, heated debates exist in the community comparing these two types of algorithms, mostly from an empirical perspective. To approach the above question, in this paper, we revisit scalarization from a theoretical perspective. We focus on linear MTL models and study whether scalarization is capable of fully exploring the Pareto front. Our findings reveal that, in contrast to recent works that claimed empirical advantages of scalarization, scalarization is inherently incapable of full exploration, especially for those Pareto optimal solutions that strike the balanced trade-offs between multiple tasks. More concretely, when the model is under-parametrized, we reveal a multi-surface structure of the feasible region and identify necessary and sufficient conditions for full exploration. This leads to the conclusion that scalarization is in general incapable of tracing out the Pareto front. Our theoretical results partially answer the open questions in Xin et al. (2021), and provide a more intuitive explanation on why scalarization fails beyond non-convexity. We additionally perform experiments on a real-world dataset using both scalarization and state-of-the-art SMTOs. The experimental results not only corroborate our theoretical findings, but also unveil the potential of SMTOs in finding balanced solutions, which cannot be achieved by scalarization., Comment: Accepted at NeurIPS 2023

We study the problem of allocating a set of indivisible chores to three agents, among whom two have additive cost functions, in a fair manner. Two fairness notions under consideration are envy-freeness up to any chore (EFX) and a relaxed notion, namely envy-freeness up to transferring any chore (tEFX). In contrast to the case of goods, the case of chores remain relatively unexplored. In particular, our results constructively prove the existence of a tEFX allocation for three agents if two of them have additive cost functions and the ratio of their highest and lowest costs is bounded by two. In addition, if those two cost functions have identical ordering (IDO) on the costs of chores, then an EFX allocation exists even if the condition on the ratio bound is slightly relaxed. Throughout our entire framework, the third agent is unrestricted besides having a monotone cost function.

To mitigate the bias exhibited by machine learning models, fairness criteria can be integrated into the training process to ensure fair treatment across all demographics, but it often comes at the expense of model performance. Understanding such tradeoffs, therefore, underlies the design of fair algorithms. To this end, this paper provides a complete characterization of the inherent tradeoff of demographic parity on classification problems, under the most general multi-group, multi-class, and noisy setting. Specifically, we show that the minimum error rate achievable by randomized and attribute-aware fair classifiers is given by the optimal value of a Wasserstein-barycenter problem. On the practical side, our findings lead to a simple post-processing algorithm that derives fair classifiers from score functions, which yields the optimal fair classifier when the score is Bayes optimal. We provide suboptimality analysis and sample complexity for our algorithm, and demonstrate its effectiveness on benchmark datasets., Comment: ICML 2023. Code is at https://github.com/rxian/fair-classification. Comparison to v2: corrected proof of Theorem 4.4

Federated learning (FL) allows agents to jointly train a global model without sharing their local data. However, due to the heterogeneous nature of local data, it is challenging to optimize or even define fairness of the trained global model for the agents. For instance, existing work usually considers accuracy equity as fairness for different agents in FL, which is limited, especially under the heterogeneous setting, since it is intuitively "unfair" to enforce agents with high-quality data to achieve similar accuracy to those who contribute low-quality data, which may discourage the agents from participating in FL. In this work, we propose a formal FL fairness definition, fairness via agent-awareness (FAA), which takes different contributions of heterogeneous agents into account. Under FAA, the performance of agents with high-quality data will not be sacrificed just due to the existence of large amounts of agents with low-quality data. In addition, we propose a fair FL training algorithm based on agent clustering (FOCUS) to achieve fairness in FL measured by FAA. Theoretically, we prove the convergence and optimality of FOCUS under mild conditions for linear and general convex loss functions with bounded smoothness. We also prove that FOCUS always achieves higher fairness in terms of FAA compared with standard FedAvg under both linear and general convex loss functions. Empirically, we show that on four FL datasets, including synthetic data, images, and texts, FOCUS achieves significantly higher fairness in terms of FAA while maintaining competitive prediction accuracy compared with FedAvg and state-of-the-art fair FL algorithms.

A wide range of resources have been developed to support the learning of high frequency words and academic vocabulary. In contrast, technical vocabulary, which can account for as much as 30% of the words in a technical text (Chung & Nation, 2004), has received less attention from language practitioners so far. Our project aims at developing resources to assist science and engineering undergraduate students in the learning and use of technical vocabulary. This presentation will discuss data collected by survey and focus groups and explain the principles of resource development. Sample learning objects will be demonstrated

A wide range of resources have been developed to support the learning of high frequency words and academic vocabulary. In contrast, technical vocabulary, which can account for as much as 30% of the words in a technical text (Chung & Nation, 2004), has received less attention from language practitioners so far. Our project aims at developing resources to assist science and engineering undergraduate students in the learning and use of technical vocabulary. This presentation will discuss data collected by survey and focus groups and explain the principles of resource development. Sample learning objects will be demonstrated

A wide range of resources have been developed to support the learning of high frequency words and academic vocabulary. In contrast, technical vocabulary, which can account for as much as 30% of the words in a technical text (Chung & Nation, 2004), has received less attention from language practitioners so far. Our project aims at developing resources to assist science and engineering undergraduate students in the learning and use of technical vocabulary. This presentation will discuss data collected by survey and focus groups and explain the principles of resource development. Sample learning objects will be demonstrated

This paper examines the impact of discipline-specific language instruction on student writing in an English for Academic Purposes (EAP) course offered to civil engineering undergraduate students at a university in Hong Kong. Textual analysis of student scripts and student interviews were used to investigate the effectiveness of the course in helping students to write better Introduction and Literature Review sections for their Final Year Project reports. The great majority of students were able to produce effective texts which conform to the discourse conventions of academic writing in civil engineering. However it is difficult to separate the impact of language instruction on student writing from other influential factors. Two major pedagogical implications are drawn: the need for more scaffolding of input tasks and for more teaching of academic reading skills. © 2018, The Asian Journal of Applied Linguistics.

This paper examines the impact of discipline-specific language instruction on student writing in an English for Academic Purposes (EAP) course offered to civil engineering undergraduate students at a university in Hong Kong. Textual analysis of student scripts and student interviews were used to investigate the effectiveness of the course in helping students to write better Introduction and Literature Review sections for their Final Year Project reports. The great majority of students were able to produce effective texts which conform to the discourse conventions of academic writing in civil engineering. However it is difficult to separate the impact of language instruction on student writing from other influential factors. Two major pedagogical implications are drawn: the need for more scaffolding of input tasks and for more teaching of academic reading skills. © 2018, The Asian Journal of Applied Linguistics.

This paper examines the impact of discipline-specific language instruction on student writing in an English for Academic Purposes (EAP) course offered to civil engineering undergraduate students at a university in Hong Kong. Textual analysis of student scripts and student interviews were used to investigate the effectiveness of the course in helping students to write better Introduction and Literature Review sections for their Final Year Project reports. The great majority of students were able to produce effective texts which conform to the discourse conventions of academic writing in civil engineering. However it is difficult to separate the impact of language instruction on student writing from other influential factors. Two major pedagogical implications are drawn: the need for more scaffolding of input tasks and for more teaching of academic reading skills. © 2018, The Asian Journal of Applied Linguistics.

Students at different levels of education are often required to write analytically. At middle or high school, they may need to write (for example) an analysis of the causes of an accident. At university, they may need to write analytical essays to examine the relationship between two concepts or ideas. However, many students find writing analytically a daunting task. They tend to be more familiar with descriptive writing, having learned how to write stories for many years, and may be at a loss about what to do when they are asked to produce a familiar text type. Some weaker students may even fail the task simply because they lack the necessary writing skills. To help them solve this problem, I have developed some tasks which aim to prepare students to write an analytical report about a disaster. I have used the materials with my university classes, with positive response. I will now describe the procedure for each task and outline the rationale for the design.

Students at different levels of education are often required to write analytically. At middle or high school, they may need to write (for example) an analysis of the causes of an accident. At university, they may need to write analytical essays to examine the relationship between two concepts or ideas. However, many students find writing analytically a daunting task. They tend to be more familiar with descriptive writing, having learned how to write stories for many years, and may be at a loss about what to do when they are asked to produce a familiar text type. Some weaker students may even fail the task simply because they lack the necessary writing skills. To help them solve this problem, I have developed some tasks which aim to prepare students to write an analytical report about a disaster. I have used the materials with my university classes, with positive response. I will now describe the procedure for each task and outline the rationale for the design.

Students at different levels of education are often required to write analytically. At middle or high school, they may need to write (for example) an analysis of the causes of an accident. At university, they may need to write analytical essays to examine the relationship between two concepts or ideas. However, many students find writing analytically a daunting task. They tend to be more familiar with descriptive writing, having learned how to write stories for many years, and may be at a loss about what to do when they are asked to produce a familiar text type. Some weaker students may even fail the task simply because they lack the necessary writing skills. To help them solve this problem, I have developed some tasks which aim to prepare students to write an analytical report about a disaster. I have used the materials with my university classes, with positive response. I will now describe the procedure for each task and outline the rationale for the design.

This paper reports the first phase of an Action Learning Project. It covers the research questions, the design and objectives of the course, the modes of training, the evaluative measures used, the results and implications for improvements. The findings will be used to improve the course (including better measures) to determine its effectiveness in the second phase of this project.

This paper reports the first phase of an Action Learning Project. It covers the research questions, the design and objectives of the course, the modes of training, the evaluative measures used, the results and implications for improvements. The findings will be used to improve the course (including better measures) to determine its effectiveness in the second phase of this project.

This paper reports the first phase of an Action Learning Project. It covers the research questions, the design and objectives of the course, the modes of training, the evaluative measures used, the results and implications for improvements. The findings will be used to improve the course (including better measures) to determine its effectiveness in the second phase of this project.