Department of Systems Engineering & Engineering Management

The following describes in more details a number of sampled research topics being actively pursued in the department. The names of the investigator/contact persons are also given. You are welcome to contact them for more details regarding particular topics of your interest.

Algorithmic Studies for Time-Varing Network Models
To investigate network models in which the structure/parameters of the network under consideration are changing dynamically and/or stochastically. Algorithms to find optimal/approximate solutions will be the focus of the study. (X. Cai)

Audio-Video Information Retrieval and Translingual Speech Retrieval
This project integrates Chinese speech recognition and information retrieval technologies for speech retrieval, based on a large archive of television news broadcasts. We also incorporate the use of video information to enhance speech retrieval accuracy. The project has recently been extended into a translingual retrieval context, where English textual queries are used to retrieve Mandarin audio documents. (H.Meng)

Automated Learning and Reasoning
To explore automated learning methodologies suitable in extracting useful knowledge and patterns from historical data, transform it into suitable representation in reasoning systems to aid decision making. (W.Lam)

Automatic Construction of Parallel and Comparable Corpus
To mine the parallel and comparable corpus from the World Wide Web automatically. In the World Wide Web, information in more than one languages are available. In this project, we shall develop techniques to construct the multilingual corpus based on the properties of parallelism and comparability. (C.Yang)

Automatic Cross-Lingual Thesaurus
To construct a cross-lingual thesaurus automatically based on the statistical information from parallel corpus and comparable corpus. Artificial intelligence techniques, such as neural network and constraint network, will be utilized. (C.Yang)

Automatic Text Summarization
To develop an automatic technique for text summarization based on fractal theory. (C.Yang)

Berth Allocation of Container Terminals
Container terminals form an important link in the transport chain for transshipment of cargo containers. Berth allocation aims to determine the optimal berthing times and positions of vessels in a container terminal, so as to reduce the service time of vessels and increase the resource utilization of berths. We will study the problem under a new framework scheduling models. (X.Cai)

Chinese Database and Information Retrieval
To design and develop a database and an intelligent information retrieval systems with user friendly interface in Chinese, especially in a form of natural language. (W.Lam, K.F.Wong)

Conically Ordered Optimization, with Applications
In this project we study optimization models in which the conic “nonnegativity” is used to model the objective preference, as well as in the constraints. Immediate applications of such models include multiple objective programming and conic optimization. Theory and solution models will be developed for such models. Engineering applications will be considered. (S.Zhang)

Content-Based Color Image Retrieval
To develop browsing and searching techniques for large scale color image databases based on chromatic, texture and shape features. Neural networks are adopted for automatic generation of the image categories for browsing. Relevance feedback will be utilized to enhance the queries. (C.C.Yang)

Continuous Basel II Requirements Compliance System (CBRCS)
Over the past few years, risk assessment, risk management, and credit rating have become extremely important. After joining WTO, banks and firms in Hong Kong SAR and China required to follow the new regulations, such as, the Basel II Accord. Potential projects include development of new risk assessment and credit rating procedures, development of new systems to support preparation of data to meet the new requirements (such process is defined to be Extraction, Transformation, and Loading (ETL)), etc. This project is a collaboration between TechnoSolve, a subsidiary of Lippo Group, and CUHK. Which develops a risk assessment and management system for banking industry in both Hong Kong and China to meet the new requirements from Basel Committee called Continuous Basel II Requirements Compliance System (CBRCS). (J.Yen)

Continuous Financial Information Disclosure and e-Auditing System
China is facing a quite different set of issues in credit rating and market monitoring, for example, due diligence and data quality. Be able to help investors to identify abnormal or erroneous data on the financial documents is extremely important and challenging and this is the major focus of our research. The second project we have is to develop the Continuous Financial Information Disclosure and e-Auditing System to solve such problem by improving the transparency and efficiency of listed forms in China . Such system monitors the transactions captured in databases of banks and enterprises to generate continuous financial statements. The under laying technologies included XML, XBRL, software agents, data mining, and statistical analysis. (J.Yen)

Continuous-Time Portfolio Selection
To investigate continuous-time optimal portfolios for a dynamic market with consumptions and transaction costs. (X.Zhou)

Coordination of a Supply Chain with Risk-Averse Agents
The extant supply chain management literature has not addressed the issue of coordination in supply chains involving risk-averse agents. In this project, we take up this issue and begin with defining a coordinating contract as one that results in a Pareto-optimal solution acceptable to each agent. We develop coordinating contracts for following cases: (i) the supplier is risk neutral and the retailer maximizes his expected profit subject to a downside risk constraint, (ii) the supplier and the retailer each maximize his own mean-variance trade-off, and (iii) the supplier and the retailers each maximize his own expected utility. In each case, we show how we can find the set of Pareto-optimal solutions, and then design a contract to achieve the solutions. We also exhibit a case in which we obtain Pareto-optimal sharing rules explicitly, and outline procedure to obtain Pareto-optimal solutions. (H.Yan)

Crime Analysis and Management System
To apply knowledge engineering, map-based and AI techniques for development of decision aids to fight crimes; to employ data mining and information extraction techniques for identification of meaningful patterns; and to use pattern matching techniques for facial composition. (K.P.Lam)

Cross-Lingual Information Retrieval
To design different techniques and algorithms to facilitate access of English information using Chinese queries, and vice versa, and to develop a system base on them. (K.F.Wong, W.Lam)

Cross-lingual Information Retrieval
To develop an information retrieval system to cross the language boundary. In this project, we shall focus on the corpus-based approach to overcome the weakness of the dictionary-based approach. (C.Yang)

Data Mining and Web Mining
Data mining is a powerful technology being widely adopted to help decision makers focus on the most important nontrivial/predictive information/patterns that can be extracted from large amounts of data they continuously accumulate in their daily business operations or obtain from datafeeds or WWW on the Internet. This project focuses on mining patterns in a large dataset including association rule mining, text mining, Web mining, etc. (X.Yu)

Data Stream Processing
Data streams arise naturally in many real-world applications like stock market. Stock market traders need to monitor tens of thousands of data coming as streams in an on-line fashion, and need to analyze and make decisions based on the data streams they have received as soon as they can. For instance, a stock market trader may need software tools to spot arbitrage opportunities. The data streams are open-end in nature, and are huge in volume. It is impossible to hold such whole data streams in memory to analyze, and is difficult to store such data streams on disk for fast processing. This fact imposes high demands on processing the data streams by reducing the number of data scans down to one as the so-called “you only got one look” with bounded memory space. This project focuses on data stream processing/mining techniques classification and text classification, etc. (X.Yu)

Dividend Distributions and Risk Controls
To explore a new area on risk and dividends distribution management for insurance companies by using techniques in stochastic optimal control. (X.Zhou)

Dual Control
Except for a few ideal situations, an optimal control usually pursues two often conflicting objectives: To drive the system toward a desired state, and to perform active learning to reduce the systems uncertainty. The dual roles of an optimal control, optimization and estimation, in general situations, cannot be separated. This coupling between optimization and estimation, in general situations, cannot be separated. This coupling between optimization and estimation makes an analytical form of optimal control, in most situations, unattainable. The research goal is to develop some embedding schemes in order to achieve optimal control laws with an active learning property for certain classes of dual control problems. (D.Li)

Dynamic Portfolio Selection with a Mean-Variance Formulation
The research goal is to seek optimal investment strategies for dynamic portfolio selection problems with a mean-variance formulation. To seek an optimal dynamic portfolio policy within a mean-variance framework implies to achieve a dual balance between the expected return and the risk and between the short term and long term benefits. Variance minimization is a notorious problem in stochastic control due to its associated property of nonseparability. Separation schemes can be developed to overcome this difficulty of nonseparability. Further research efforts are needed to improve the portfolio selection models and to derive full feedback optimal investment policies. (D.Li)

Dynamic Pricing and Risk Evaluation
Two problems bring up this topic. One is the dynamic pricing for the natural gas pipelines and for the ship transportation for oil. The other is the risk evaluation in the multi-stage supply chains. Synthesizing the risk analysis and optimization or developing an efficient procedure for the value-at-risk (VaR) should be considered for various products across industries. (Y.Feng)

Expert Systems and Knowledge Discovery
To investigate the theory and techniques for to partial or full automation of the the time-consuming process of expert knowledge elicitation through automatic knowledge discovery or learning from data. (W.Lam)

FPGA Implementation for Bioinformatics
Various algorithms on Hebbian learning, inference network, transitive closure, and phylogenetic tree mapping of DNA sequences are under investigation for efficient implementation on FPGA (field programmable gate array), including an embedded platform of Xilinx Virtex II Prof with PowerPC microprocessors. (K.P.Lam)

Hidden Convexity
The research goal is to develop sufficient conditions to identity hidden convex minimization problems. A nonconvex minimization problem is called a hidden convex minimization problem if there exists an equivalent transformation such that the transformed minimization problem is convex. Sufficient conditions that are independent of transformations can be derived for identifying such class of seemingly nonconvex minimization problems that are equivalent to convex minimization problems. A global optimality can be thus achieved for this class of hidden convex optimization problems by suing local search methods. (D.Li)

Incentives for Advance Ordering in a Supply Chain
To address the issues in designing incentives for advance ordering in a supply chain. In particular, we investigate what incentives the firms in the higher echelon of a supply chain can employ to induce the firms in the lower echelon to place advance orders and how each of these incentives affect the ordering behaviour of the firms in the lower echelon. (Y.Chen)

Inference Machines
To design a real-time systems for control and optimization applications which involve hardware prototyping and software simulation of inference networks. (K.P.Lam)

Intelligent Information Retrieval
To investigate intelligent retrieval models and incorporation of feedback information during the retrieval process. (W.Lam)

Internet Performance Analysis and Optimization
To analyze the Internet traffic, focusing on dependence, heavy-tailed distributions, and rare events; and to apply dynamic resource management (e.g., scheduling, caching, load balancing, flow control), to achieve optimal resource utilization and quality of service. (D.Yao)

Intraday Financial Prediction
Linear and nonlinear modeling techniques are applied for financial prediction using intraday data. Computational intelligence methodologies including fuzzy expect system, neural network, and multiple models are developed for receding-horizon prediction that makes effective use of t he timely arrival of high frequency intraday news. (K.P.Lam)

Inventory Strategies of Online Retailer
To examine the inventory strategies of Internet retailers who face options of holding own inventories, outsourcing through the third party(ies), or a combination. (Y.Chen, Y.Feng)

Logic in Information Retrieval
To characterize different information retrieval models in logic using the Aboutness Theorem. A set of axioms is being defined to represent the functionality of information retrieval. These axioms would then be used as a functional benchmark suite. (C.H.Cheng, K.F.Wong)

Manufacturing Systems and Production Planning
To develop techniques for designing reasonably good and easy-to-implement policies for large and complex manufacturing systems. (X.Zhou)

Map-based Delivery Route Planning with Dynamic Updates
To develop a map-based system for route and load planning for delivery, suitable for use by small- and medium-sized enterprises (SMEs) in Hong Kong. Algorithms for dynamic re-routing based on traffic conditions will be investigated. (C.H.Cheng and J.M.Y.Leung)

Multi-echelon Supply Chain Management
To optimize the inventory of a multi-echelon supply chain using constraint network and relative sensitive algorithm. (C.Yang)

New Scheduling Models with Negotiable Third-Party Machines
To study models and algorithms for a branch of scheduling problems in which third-party machines may have to be secured to process certain jobs, through negotiation. Factors such as the due dates to process the jobs as required by the relevant contracts, and the price and logistic costs to use third-party machines, will be addressed in the scheduling models. (X. Cai)

Nonconvex Optimization and Global Optimization
The research goal is to develop equivalent transformations for generating a saddle point for nonconvex optimization problems. A saddle point condition is a sufficient condition for optimality. A saddle point can be generated in an equivalent representation space for nonconvex optimization problems that do not have a saddle point in their original settings. Certain equivalent transformations may convexify the perturbation function and a zero duality gap can be thus achieved. This investigation would lead to some efficient dual search algorithms that ensure the global optimality for a class of nonconvex optimization problems. (D.Li)

Nonlinear –1 Programming with Applications to Financial Engineering
We study the application of the SDP relaxation and the randomization methods for a certain class of p-1 programming problems. As an application of such models, we study the asset picking problem, which plays an important role in the portfolio theory and financial engineering. (S.Zhang)

Nonlinear Integer Programming
The research goal is to establish convergent duality theory and to develop efficient solution algorithms for large-scale nonlinear integer programming problems. The fundamental target underlying our theoretical development is to eliminate duality gap in the classical Lagrangian dual formulation. We have developed nonlinear Lagrangian theory that has yielded several new dual formulations with asymptotic zero duality gap. The key concept is the construction of a nonlinear support for a nonconvex piecewise-constant perturbation function. Our numerical implementation of a duality-gap reduction process relies on some novel “cutting” procedures. Performing objective-level cut, objective contour cut or domain cut reshapes the perturbation function, thus exposing eventually an optimal solution to the convex hull of a revised perturbation function and guaranteeing a zero duality gap for a convergent Lagrangian method. Applications include nonlinear knapsack problems, constrained redundancy optimization in reliability networks, and optimal control problems with integer constraints. (D.Li)

Optimal Decisions for Production and Delivery of Perishable Products
We will study a situation where a variety of perishable products are to be produced and delivered to an export market, using a transportation service subject to stochastic disruptions. Optimal decisions are to be sought via stochastic optimization, by taking into account information on the possible delivery time of the transporter and the random decaying process of the perishable products. Topics to be addressed include those on modeling, propositions of optimal policies, and algorithms. (X. Cai)

Optimal Dynamic Portfolio Selection with Mean-Variance Formulation
To develop efficient solution algorithms to derive an analytical solution for multiple period portfolio selection. (D.Li, X.Zhou)

Optimal Selling Mode for Perishable Items: Posted Price and/or On-line Auction?
To study the optimal switching point in selling perishable items. Typically, those items are first sold by the posted price mode, and then after a certain point in time sold through online auction. When is the optimal stopping (start) time for posted price selling (auctioning)? (Y.Chen, Y.Feng)

Optimization of Stochastic Dynamic Systems
To investigate optimization problems for two types of stochastic dynamic systems: systems in which white noises exist in the dynamics and systems that are driven by discrete events. (X.Zhou)

Portfolio Optimization under Minimax Measures
Portfolio selection problems under a kind of new risk measures will be studied. Issues on model analyses, solution examination, and algorithms will be investigated. (X. Cai)

Pricing and Inventory Management for Perishable Products
Selling a fix number of perishable products need differentiate prices and segment customers properly. The focus is centered on optimal pricing and inventory control policies for airline seats, hotel rooms and energy products. (Y. Feng)

Risk Analysis of Inventory Models with Forecast Updates
The standard treatment of supply chain models with demand forecasting confine themselves to the optimization of the expected value of a given cost or profit measure without consideration of a risk measure. Due to highly uncertain market demand, the use of the expected objective measure may be not justified. To reduce the impact of the high uncertainty of market demand, the advanced reservation and the general mean-variance decision framework are introduced. We show that a mean-variance trade-off analysis with advanced reservation can be carried out efficiently, and how the corresponding inventory decisions differ from the traditional decisions. (H.Yan)

Robust Optimization
To study the theory and applications of robust optimization. The issue of robust optimization arises when the data for an optimization model is uncertain. The problem of this type is pervasive in engineering design and management, where the measurements can at best be done only approximately. (S.Zhang)

Semidefinite Programming and Applications
To study advanced solution methods such as primal-dual interior point method in solving semidefinite programming problem and related modelling issues. (S.Zhang)

Stochastic Linear-quadratic Control with Financial Applications
To investigate stochastic linear-quadratic control problems with indefinite cost weighting matrices, using modern optimization techniques such as semidefinite programming as computational means. Applications in portfolio selection and financial index tracking will be studied. (D.Yao, S.Zhang, X.Zhou)

Stochastic Programming
To investigate high performance optimization techniques in stochastic programming with applications in financial problems. (S.Zhang)

Supply Chain Logistics
To develop methods and technologies that support decision making in supply chain logistics. Two broad issues will be addressed: (a) modelling, analysis and optimal design of a supply (b) coordinating quality control schemes and production/replenishment decisions in multi-stage systems. (D.Yao)

Temporal Information Extraction
To research into the properties of temporal information embedded in Chinese text and to design novice algorithms to extract such properties for the purpose of conceptual information retrieval. (K.F.Wong)

The "Author Once, Present Anywhere" (AOPA) Software Platform
The AOPA Software Platform aims to help web designers to achieve undersal accessibility, i.e. maintain a single content repository and the content will automatically adopt usability-optimized presentation styles to cater for various client devices, such as displayless voice browsers, mobile mini-browsers and regular Web browsers. AOPA utilizes W3C standard technologies such as XML and XSLT. It alsso incorporates core technologies developed in the Human-Computer Communications Laboratory, including Cantonese speech recognition and speech synthesis. (H.Meng)

Towards Multi-modal Human-Computer Dialog Interactions with Minimally Intrusive Biometric Security Functions
This project aims to develop human-centric interface technologies to support secure computing by a diversity of users in a variety of usage contexts. Human-centric interface technologies embrace the user's natural communicative modalities at the center of human-computer interaction (HCI). For example, the user can speak and point to the computer, and the computer can "see" the use's face through the camera, "Hear" the user's spoken commands via handsets and microphones, and "sense" the user's touch during pointing with his fingers/stylus pens/mice. As computing permeates our everyday lives, security to computers, networks and content becomes and issue of prime importance. User authentication should be transparent to the human-computer interaction. In this project, we will focus on a speak-and-point interface secured with facial identification, speaker and fingerprint authentication on a mobile platform. (H.Meng)

Traffic Inference Engine
To develop models and methods for transit-time estimation of a road network, based on information on a collection of trips between various origins and destinations in the network. Issues of model accuracy and robustness, data updating and filtering will be investigated. (J.M.Y.Leung and C.H.Cheng)

XML Database and Query Processing
XML is emerging as a major standard for representing data on the World-Wide-Web. This project focuses on XML database issues including XML storage models, query processing, query optimization, indexing, and updating issues. (X.Yu)