Title: Machine Learning: From Concepts to Deployment: A Perspective of Granular Computing Witold Pedrycz Professor, University of Alberta, Canada URL: http://www.ece.ualberta.ca/~pedrycz/ Talk Abstract: Over the recent years, we have been witnessing spectacular achievements of Machine Learning (ML) with highly visible accomplishments encountered, in particular, in natural language processing and computer vision impacting numerous areas of human endeavours. Driven inherently by the technologically advanced learning and architectural developments, Machine Learning constructs are highly impactful coming with far reaching consequences; just to mention autonomous vehicles, control, health care imaging, decision-making in critical areas, among others. The objective of this talk is to elaborate on the key technical and non-technical issues that have surfaced in the course of the developments of new architectures of ML and expanding a spectrum of their applications. Whereas the technical stumbling blocks are relatively well identified and intensively researched, there is a sphere of equally important non-technical quests. We focus our discussion on identifying, formulating and discussing the main challenges including (i) brittleness, explainability and interpretability, (ii) credibility and self-awareness of results generated by ML constructs, (iii) privacy, and (iv) sustainable computing. We demonstrate which key pursuits have been put forward on the agenda of interpretable AI (XAI) and have made their way to the list of goals of Green AI. Furthermore, we discuss a promising direction of knowledge-directed (informed) ML. Throughout the talk, it is advocated that Granular Computing offers a conceptually and algorithmically sound framework aimed at delivering some solutions to the above stated changes.
	Title: 基于物理信息时空机器学习的智能油藏模拟 Zhangxing Chen Professor,宁波东方理工大学 / University of Calgary URL:  https://schulich.ucalgary.ca/contacts/zhangxing-john-chen Talk Abstract: 机器学习作为一种实现人工智能的方法，其高效的运算能力及强大的预测功能在近几年中获得广泛关注，并被大量应用于生活中的各个方面：如自然语言处理、金融行业、互联网等领域。但机器学习方法在油气行业中的应用仍面临诸多挑战。本报告将介绍结合物理和数据驱动的机器学习方法在油气行业中应用进展和前景，重点介绍油藏智能模拟及其代理模型: • 基于全连接神经网络代理模型; • 基于卷积编码器解码器代理模型; • 基于时空网络的代理模型; • 实例分析: 油气藏数值模拟智能加速，产能智能预测，压裂液返排智能预测等。
	Title: Meta-learning with Many Tasks James Tin-yau KWOK (郭天佑) Department of Computer Science and Engineering, The Hong Kong University of Science and Technology URL:   https://home.cse.ust.hk/~jamesk/ Talk Abstract: In many machine learning applications, one only has a limited number of training samples. To alleviate this problem, a successful approach is meta-learning, which tries to extract meta-knowledge from similar historical tasks. Obviously, the larger the number of tasks to learn from, the more meta-knowledge can be learned. However, popular meta-learning algorithms like MAML only learn a globally-shared meta-model. This can be problematic when the task environment is complex, and a single meta-model is not sufficient to capture diversity of the meta-knowledge. Moreover, the sampling of tasks in each iteration also increases variance in the stochastic gradient, resulting in slow convergence. In this talk, we propose to address these problems by (i) using multiple meta-models for initialization, and (ii) incorporate variance reduction into meta-learning for faster convergence. Experiments on various meta-learning tasks demonstrate its effectiveness over state-of-the-art algorithms. Bio: Prof. Kwok is a Professor in the Department of Computer Science and Engineering, Hong Kong University of Science and Technology. Prof. Kwok is serving as an Associate Editor for the IEEE Transactions on Neural Networks and Learning Systems, Neurocomputing, Artificial Intelligence Journal, International Journal of Data Science and Analytics, and on the Editorial Board of Machine Learning. He is also serving as Senior Area Chairs of major machine learning / AI conferences including NeurIPS, ICML, ICLR, IJCAI. Prof Kwok will be the IJCAI-2025 Program Chair.
	Title: Evolutionary Computing and Complex Networks for Metabolomics and Precision Medicine Ting Hu Professor, Department of Computer Science, Memorial University, Canada URL:  https://www.cs.queensu.ca/people/Ting/Hu Talk Abstract: Machine learning has the remarkable capability to uncover intricate patterns and relationships within data. However, the consequential decisions made based on these model predictions can profoundly affect human lives. As machine learning models find their way into high-stakes domains such as medicine, job hiring, and criminal justice, concerns about fairness, transparency, and accountability have rightfully emerged. In response, there is a growing need not only to create highly accurate prediction models but also to comprehend and elucidate the inner workings of these predictive systems. Evolutionary computing, a versatile meta-learning approach, can generate innovative, multi-objective, and diverse solutions to optimization and learning challenges. It holds exceptional promise in the realm of crafting solutions for Explainable AI (XAI). In this talk, we discuss the landscape of explainability and its associated concepts. We highlight the potential of evolutionary computing as a means of crafting comprehensive explanations for machine learning.
	Title: 大数据机器学习理论研究现状及其应用前景 Wensheng Zhang Professor,University of Chinese Academy of Sciences, China URL:  http://people.ucas.ac.cn/~wenshengzhang Talk Abstract: 随着社会对数据深度理解的强烈需求，大数据机器学习逐步被学术界和企业界重视，本报告主要从机器学表示理论和实现技巧角度讲解如何理解机器学习、解析机器学习基石、探索大数据机器学从理论和算法途径。重点聚焦在深入解析机器学习的本质，探索大数据催生机器学习理论和技术的变革，提供从传统机器学习到大数据机器学习转换的思路，通过近期人工智能创新应用的典型案例分析大数据机器学习理论研究现状，展望未来产业应用前景。
	Title: Imbalanced Data Learning: Always Needed? Yiu-ming Cheung Professor, Hong Kong Baptist University, China URL:  https://www.scholat.com/xmzhang8.cn Talk Abstract: In many practical problems like disease diagnosis and anomaly detection, the number of data forming difference classes can be quite imbalanced, which could make the performance of the most machine learning methods become deteriorate to a certain degree. In particular, although deep learning has made great progress, a good model often requires a large amount of artificially balanced and annotated data. Unfortunately, real-world data are often unbalanced, typically exhibiting a long-tailed distribution, which refers to a small number of classes with abundant training samples but the remaining large number of classes only with very few training instances. Under the circumstances, the performance of deep learning models trained on long-tailed data declines sharply in the tail classes. In this talk, we will first introduce the class imbalanced learning and its related techniques. Then, whereas the class imbalance is quite common from the practical perspective, we will attempt to address the problem naturally arisen: When should such a class imbalance be taken into account in the machine learning tasks? Furthermore, the impact of long-tailed data on deep learning models will be analyzed. The research progress in this area will be reviewed, including some representative methods in the literature. Besides, our recent work for long-tail learning will be introduced. Lastly, the potential research directions in this field will be discussed.