Research Project

Gene Analysis

This project is based on single cell RNA-Seq data from bladder cancer patients to identify subtypes of bladder cancer and to discover some key pathogenic genes.

Prediction of Recurrence of Kidney-Stone

This project uses big data technology to predict the recurrence probability of kidney stones and the possible recurrence time by case analysis.

 

Audio Dialogue Collection

In this project, an audio acquisition system was designed to collect the dialogue between doctors and patients. This lays the foundation for medical Natural Language Processing work.

 

Patient visiting system

This project develops intelligent patient visit system and data acquisition system by cooperating with our local hospital. The system is closely integrated with the needs of the doctors. It uses advanced Internet of things technology to achieve large-scale, all-round patient and medical data collection, and provides services for patients and doctors through large data processing technology.

 

 

Learning Analytics

This project uses the latest machine learning and big data methods, such as latent semantic model and deep learning, to explore the learning data (desensitization) of the Chinese University Hong Kong (Shenzhen), in order to help to build a smart campus.

 

PNM intelligent analysis algorithm PNM

The project cooperates with HUAWEI Technologies Co. Ltd to design and develop PNM intelligent analysis algorithm, including: (1) the extraction of fault automatic recognition and quantitative characteristics of optical fiber network uplink and downlink spectrum. (2)fault feature extraction and automatic recognition and quantification of QAM constellation constellation restoration. (3) optical fiber network communication channel equalization coefficient (sequence clustering). (4) the expression of topology and fault location algorithm.

 

Phased array optimization technique

This project, in collaboration with the fourteenth Research Institute of China Electronic Technology Group Corporation, aims to study the optimization technology of phased array.

 

Human-machine dialogue system based on deep learning

The human-machine natural language dialogue system is the most challenging research in the field of artificial intelligence, carrying the dream of human science and technology. This project aims to implement a human-machine natural language question answering system in a specific field based on deep learning technology. Based on the Java platform, we have developed a demonstration prototype that integrates speech recognition, problem recognition, answer generation, and speech synthesis. According to the user's voice input, the system first identifies the user's information, then identifies the user's question, finds the answer in the system, and evaluates the answer. The system finally outputs an answer which is the most appropriate to the user's question. At present, our developing system has integrated several functions: such as free dialogue, medical question and answer, writing ancient poetry and so on.