Discover the basic concepts of cluster analysis, and then study a set of typical clustering methodologies, algorithms, and applications. This includes partitioning methods such as k-means, hierarchical methods such as BIRCH, and density-based methods such as DBSCAN/OPTICS. Moreover, learn methods for clustering validation and evaluation of clustering quality. Finally, see examples of cluster analysis in applications.
Cluster Analysis in Data Mining is third course in Coursera's new data mining specialization offered by the University of Illinois Urbana-Champaign. The course is a 4-week overview of data clustering: unsupervised learning methods that attempt to gr…
Cluster Analysis in Data Mining is third course in Coursera's new data mining specialization offered by the University of Illinois Urbana-Champaign. The course is a 4-week overview of data clustering: unsupervised learning methods that attempt to group data into clusters of related or similar observations. The course covers two most common clustering methods--K means and hierarchical clustering--as well as more than a dozen other clustering algorithms. Grading is based on 4 weekly quizzes with 3 attempts each.
Cluster Analysis is taught by Professor Jiawei Han who was the instructor for the first course in the data mining specialization: Pattern Discovery in Data Mining. The quality of the slides, instruction and organization of materials in this course is slightly better than the pattern discovery course, but that isn't saying much: it is still below Coursera's usual high standards. The course rushes from one topic to another with instruction that is mediocre at best downright confusing at its worst. That's not to say you can't learn anything from this course, but the instruction is often more of a hindrance than a help. There are occasional in-lecture quizzes, but the graded quizzes largely fail to foster any understanding of the material. An optional programming assignment was added half way through the course; in a course about data mining, programming assignments should be front and center, not added as an afterthought to quell an outcry from students.
Cluster Analysis in Data Mining is another disappointing entry in Coursera's data mining specialization. Although the course covers many different clustering methods, poor instruction makes it hard to gain a good understanding of the material unless you are extremely attentive or watch the videos several times.
I give Cluster Analysis in Data Mining 2 out of 5 stars: Poor.
I thought the class was good for someone who already knows how to apply clustering analysis to data. I have been using different clustering algorithms in the past, this class gave me a greater overview of the other clustering methods that existed that I hadn't been exposed to. I do not recommend this for someone who is new to the concept/application of clustering.
I liked the way I was able to learn more about the newest trends in clustering algorithms, but there was too much theory, and too little practice. However, it was a fun experience, but I hope in the second iteration that the ratio of the programming assignments and the theoretical descriptions of various algorithms and papers will be equal.