Advanced Data Mining, Fusion and Applications - CSC 580(Spring-2011)
CSC-580 (Graduate Elective), CSC-490c (Undergraduate Elective)
Download course flyer here.
This course offering is to enable students of any discipline to learn about the principles of Data Mining and Data Fusion and its applications in Cyber Security and other data-rich areas. The course covers the fundamentals of High Dimensional data clustering • Types of cluster analysis and applications in anomaly detection • Meta data detection • Correlation Analysis • Data conflict detection • Resolution of semantic heterogeneity towards smooth distributed data integration • Implementation of Support Vector Machines for multi-class classification and its application in intrusion detection • Application of spatio-temporal data structures for range queries for data mining applications • Intricacies of Image feature extraction for Content-Based Image Retrieval and Annotation of Images • Advanced solutions in indexing and querying large time-series.
The fundamentals in Privacy Preserving Data Mining (PPDM) in cyber-security, including privacy preserving techniques, multi-party computing (MPC), cryptography The performance evaluations of PPDM algorithm
Privacy preservation decision trees
Privacy preservation Bayesian network
Privacy preservation K-nearest neighbor
Privacy preserving k-clustering.
The emphasis on applications from fields of Cyber Security
Financial Data Analysis
PPDM in intrusion detection and network monitoring
Emerging challenges of PPDM applications in cyber-security
For further information contact Dr. Dua.
Data Mining and Knowledge Discovery (Effective Winter-2009)
CSC-493 (Undergraduate Elective) cross-listed with CSC-579 (Graduate Elective)
Understand the architecture of a typical data mining system.
Understand various descriptive data summarization foundational techniques to measure tendency, dispersion and graphical inspection of high-dimensional data.
Understand and implement data smoothening, Gaussian noise reduction and outlier identification and correction in streaming data.
Understand and implement various normalization procedures includes z-score and ratio-scaled normalizations.
Understand Fourier transformations, discrete wavelet transformation and Parseval's theorem.
Undertstand dimensionality reduction methods using Fourier transformation.
Automatically generate concept hierarchies and discretize data using entropy based discretization and intuitive partitioning.
Understand and implement algorithms for frequent itemset discover.
Differentiate between levels, dimensions, correlation or associate and dense or sparse association rules.
Define an association rule problem from variety of business and scientific problems.
Understand differences between antimonotonic, monotonic, convertible, and inconvertible constraint-based rule mining.
Understand and resolve issues with Classification and Prediction.
Understand decision-tree based classification.
Understand and employ Bayesian classification and rule-based classification.
Understand the intricacies of Associative Classification.
Understand and employ classification accuracy and error measures.