Welcome to the Data Mining Lab at Ajou university. Data sets with millions of records and thousands of fields are increasingly common in business, engineering, medicine, and the sciences. With the amount of data doubling every few years the problem of uncovering hidden patterns or extracting useful information from such data sets is becoming an important practical issue. Research on this topic focuses on key questions such as how can one build useful models which both allow us to make predictions and also aid us to figure out the underlying process of the data generation. Research projects in our lab use theories and techniques from the intersection of computer science, statistics, and mathematics, including foundational ideas from algorithmsartificial intelligencemultivariate data analysis, Bayesian estimation, and computational statistics (from statistics), and optimization and probability theory (from mathematics). Machine learning and pattern recognition, in particular, are central to our research, providing both a sound theoretical basis and a practical framework for developing useful data analysis algorithms.  Research activities in our lab range across areas as different as hospital fraud detection, direct marketing in CRM, oil price prediction, protein function prediction in bioinformatics, etc. We hope you find our web-site useful and encourage you to explore its contents (publications, courses, seminars, and other information).

 

1. Semi-Supervised Learning
-Graph-based SSL, Transductive Learning
 

2. Kernel Methods
-Support Vector Machines(SVM)
Kernel Principal Component Analysis (KPCA)
Independent Component Analysis (ICA), etc.
 

3. Connectionist Methods
-Feed-Forward Neural Networks
Self-Organizing Map (SOM), etc.
 

4. Ensemble Learning
-Bagging
Boosting
Various Ensemble Networks of NN or DT, etc.

Mingoo Kim, Hyunjung Shin, Taesu Chung, Je-Gun Joung and JuHan Kim
Extracting Regulatory Modules from Gene Expression Data by Sequential Pattern Mining
BMC Genomics, 2011, Vol. 0, No. 0, pp. 0~ 0

Young Soo Song, Chan Hee Park, Hee-Joon Chung, Hyunjung Shin, Jihun Kim, and Ju Han Kim
Semantically enabled and statistically supported biological hypothesis testing with tissue microarr
BMC Bioinformatics, 2011, Vol. 12, No. 0, pp. 51~ 62

Tianya Hou, Hyunjung Shin, Kanghee Park, Chan-Kyoo Park and Sunghee Choi
A Semi-Supervised Learning Approach for Oil Price Prediction
Energy Economics(submitted), 2010, Vol. 333, No. 0, pp. 0~ 0

Inae Choi, Kanghee Park, and Hyunjung Shin
Sharpened Graph Ensemble for Semi-Supervised Learning
Pattern Analysis and Applications(submitted), 2010, Vol. 0, No. 0, pp. 0~ 0

Andrea Martin Lisewski, Hyunjung Shin, and Olivier Lichtarge
Frustration In Network-Based Protein Function Prediction
Science (submitted), 2008, Vol. 911, No. 0, pp. 0~ 0

Mingoo Kim, Hyunjung Shin, Taesu Chung, and JuHan Kim
Regulatory Module Identification by Sequential Pattern Mining
BMC Bioinformatics (submitted), 2007, Vol. 0, No. 0, pp. 0~ 0

Hyunjung Shin and Sungzoon Cho
Invariance of Neighborhood Relation under Input Space to Feature Space Mapping
Pattern Recognition Letters, 2005, Vol. 26, No. 6, pp. 707~ 718

Amna Ali, Hyunjung Shin, Yeolwoo An, Dokyoon Kim, Kanghee Park and Minkoo Kim
A Robust Predictive Model for Breast Cancer Survivability
Artificial Inteligence in Medicine(submitted), 2010, Vol. 447, No. 0, pp. 0~ 0

Hyungjoo Lee, Sungjoo Lee, Byungun Yoon, Hyunjung Shin
Technology clustering based on evolutionary patterns: The case of information and communications technologies
Technological Forecasting and Social Change (submitted), 2010, Vol. 0, No. 0, pp. 0~ 0

Mingoo Kim, Hyunjung Shin, Taesu Chung, and JuHan Kim
Extracting Regulatory Modules from Gene Expression Data by Sequential Pattern Mining
Bioinformatics (submitted), 2008, Vol. 0, No. 0, pp. 0~ 0