Welcome to the Spring 2021 School of Computer Science Online HDR Showcase.
This event focuses on celebrating and acknowledging the outstanding work of our Ph.D. candidates and Masters by Research students.
The event will consist of a series of online live sessions along with a set of on-demand poster presentations for you to access from the comfort of your home.
Presentations will be delivered in COVID-safe manner via Zoom. To join a live presentation, simply click on the title of the session you wish to join.
Please click on the session you wish to join.
Semi-supervised Learning, Transfer Learning, Reinforcement Learning
- Judges: Wei Liu, Ling Chen, Mukesh Prasad
- 10 AM - Event Opening
- SemiNLL: A Framework of Noisy-Label Learning by Semi-Supervised Learning
- Zhuowei Wang
- Transfer learning with multiple source domains
- Keqiuyin Li
- Hierarchical Reinforcement Learning with Optimal Level Synchronization based on a Deep Generative Model
- JaeYoon Kim
- Judges: Sanjiang Li, Nengku Yu, Yulei Sui
- Approximate Equivalence Checking of Noisy Quantum Circuits
- Xin Hong
- Quantum algorithm for graph connectivity with global queries
- Arinta Primandini Auza
- VSQL: Variational Shadow Quantum Learning for Classification
- Guangxi Li
- PACBayes Bound for Metalearning with Data Dependent Prior
- Tianyu Liu
Computer Vision, Robot
- Judges: Hua Zuo, Nico Pietroni, Nabin Sharma
- Unified Transformer Network for Object Tracking
- Fan Ma
- Removing rain from images via deep neural networks
- Ruijie Quan
- Persuasive Robots: The interplay between emotion and logic in persuasive backfiring
- Judges: Guandong Xu, Linchao Zhu, Mukesh Prasad
- Symbolic Execution in Program Analysis
- Guanqin Zhang
- Deep Learning Based Traceability Links Recovery Approach For Issue Management
- Thazin Win Win Aung
- Scaling Precision Static Taint Analysis to Industrial Enterprise Micro-services for Field-Based Sensitive Data Tracking
- Zexin Zhong
- Hierarchical topic tree: A hybrid model comprising network analysis and density
- Mengjia Wu
Machine Learning & Applications
- Judges: Angela Huo, Nabin Sharma, Marian-Andrei RIZOIU
- Evolving Gradient Boost: A Pruning Scheme Based on Loss Improvement Ratio for Learning under Concept Drift
- Kun Wang
- Learning-Based Frameworks for Automated Identifying Mental Illness Through Social Media
- Hamad Zogan
- Incident duration prediction using Machine Learning
- by Artur Grigorev
- 4:00PM - Closing Ceremony
To access this on-demand content, simply click on the title of the poster presentation you wish to watch.
A Hybrid Quantum-Classical Hamiltonian Learning Algorithm
Hamiltonian learning is central to studying complex many-body physics and the certification of quantum devices and simulators. How to learn the Hamiltonian in general with near-term quantum devices is a challenging problem. In this paper, we develop a hybrid quantum-classical Hamiltonian learning algorithm to tackle this problem. By transforming the Hamiltonian learning problem to an optimization problem using the Jaynes’ principle, we employ a gradient-descent method to give the solution and could reveal the interaction coefficients from the system’s Gibbs state measurement results. In particular, the computation of the gradients relies on the Hamiltonian spectrum and the log-partition function. Hence, as the main subroutine, we develop a variational quantum algorithm to extract the Hamiltonian spectrum and utilize convex optimization to output the log-partition function. We also apply the importance sampling technique to circumvent the resource requirements for dealing with large-scale Hamiltonians. As a proof of principle, we demonstrate the effectiveness of our algorithm by conducting numerical experiments for randomly generated Hamiltonians and many-body Hamiltonians of theoretical and practical interest.
Adaptive Graph Co-Attention Networks for Traffic Forecasting
Traffic forecasting has remained a challenging topic in the field of transportation, due to the time-varying traffic patterns and complicated spatial dependencies on road networks. To address such challenges, we propose an adaptive graph co-attention network (AGCAN) to predict traffic conditions on a road network graph.
Blockchain in 5G and 6G networks
As more and more 5G networks are deployed, the limitations of 5G networks are not only being discovered but are driving exploratory research into 6G networks as a next-generation solution. Part of these investigations includes the fundamental security and privacy problems associated with 6G technologies. Therefore, to consolidate and solidify this foundational research as a basis for future investigations, we have prepared this poster on the current state-of-play in 6G-related security and privacy. The background of 6G networks have also been presented. Moreover, this poster also separately shows what role the blockchain plays in the 5G and 6G networks. The issues and problems in blockchain network have also been introduced in this poster. What's more, the references and the future research of blockchain in 6G networks are also been detailed
Density-based detection of cell transition states to construct disparate and bifurcating trajectories
Tree- and linear-shaped cell differentiation trajectories have been widely observed in developmental biologies and can be also inferred through computational methods from single-cell RNA-sequencing datasets. A density-based trajectory inference method is introduced with the capability of constructing a diverse range of topological patterns including the most intriguing bifurcations.
Private-Encoder: Enforcing Privacy in Latent Space for Human Face Images
This poster describes a novel learning-based facial de-identification method that can semantically sanitize face image’s identity information in unconditional GAN’s latent space. The experiment results demonstrate that the proposed method achieved a balanced trade-off between image’s privacy protection and utility preservation.
TagPick : A System for Bridging Micro-Video Hashtags and E-commerce Categories
In this work, we therefore proposed a novel solution called TagPick that incorporates clues from all user behavior metadata (hashtags, interactions, multimedia information) as well as relational data (graph-based network) into a unified system to reveal the correlation between e-commerce categories and hashtags in industrial scenarios. In particular, we provide a tag-level popularity strategy to recommend the relevant hashtags for e-Commerce platform (e.g., eBay).