MCS316(Semester I 2006) Projected Clustering

Being an advanced course in Data Mining, the pre-requisite is MCS104. Other pre-requisite is a course on Algorithms.

Internal Assessment

2 Minors as per the schedule displayed on the department notice board (10 marks each)

1 Programming Assignment (10 marks) (Look here)

1 Research paper presentation  and its implementation(20 marks)(Look here)

Take Home Assignments (10 marks) (Look here)

 

No regular text will be followed. The tentative list of papers is available below.

  1. Overview of Clustering Techniques  (Ref:  Texts in MCS 104)
  2. Automatic Subspace Clustering of High Dimensional Data For Data Mining Applications (SIGMOD98)
  3. Fast Algorithms for Projected Clustering (SIGMOD99)
  4. A Monte Carlo algorithm for fast Projective Clustering
  5. HARP: A Practical Projected Clustering Algorithm
  6. A Human-Computer Interactive Method for Projected Clustering (TKDE, Apr 2004)
  7. Iterative Projected Clustering by Subspace Mining (TKDE, Feb 2005)
  8. Using emerging pattern based Projected Clustering and gene expression data for cancer detection  (Bioinformatics Conf. 2004)
  9. P3C: A Parameter less Approach to Projected Clustering (ACM KDD06)
  10. A Grid-Based Clustering Algorithm for High-Dimensional Data Streams (ADMA, 2005)
  11. A Framework for Projected Clustering of High Dimensional Data Streams

 

Useful Links for Projected Clustering

1. University of Munich

 

 

 

 

 

 

 

 

Programming Assignment I

Deadline: 30 Aug 2006

Compare  the clustering quality using 1) SSE   2)Scatter Matrix  3) Any External Criteria.

Use K-means algorithm for clustering. Prepare a 2-3 page report on comparative analysis using a dataset of your choice.

 

Top

 

 

 

 

 

Research Paper (Implementation and Presentation)

Implementation: Doc and Fast Doc Algorithms  (Deadline - 16 Oct) 

Presentation: IPCLUS ALgorithm  (Sceduled last week Oct)

Top

 

 

 

 

 

 

 

 

 

 

 

 

Take Home Assignment

 

1. Compute the complexity of the greedy algorithm for finding the superset of medoids in PROCLUS. What improvements do you suggest?  (Deadline 10 Sept)

2. Write a note on Monte-Carlo Algorithm  (approx 1500 words)  (Deadline Sept 20)

 

 

Top