Institutional Scholarship

SMOReS: Sparse Matrix Omens of Reordering Success

Show simple item record

dc.contributor.advisor Wonnacott, David G.
dc.contributor.advisor Strout, Michelle Mills Wood, Samantha 2011-10-19T12:57:13Z 2011-10-19T12:57:13Z 2011
dc.description.abstract Despite their widespread use, sparse matrix computations exhibit poor performance, due to their memory-bandwidth bound nature. Techniques have been developed that help these computations take advantage of unexploited data reuse by transforming it into data locality, generally improving performance. One such technique is to reorder the matrix prior to running a computation on it. However, reordering a matrix takes time and does not always provide performance improvements. We present a classification model that predicts, with 82% accuracy and no significantly incorrect predictions, whether reordering a matrix will improve the performance of the matrix power kernel, Akx. Our classifier is an ensemble of decision stumps generated by the AdaBoost learning algorithm and is trained on 60 matrices with a wide range of memory footprints and average number of nonzeros per row. en
dc.description.sponsorship Haverford College. Dept. of Computer Science en
dc.language.iso en_US en
dc.subject.lcsh Sparse matrices
dc.subject.lcsh Sparse matrices -- Mathematical models
dc.title SMOReS: Sparse Matrix Omens of Reordering Success en
dc.type Thesis (B.S.) en
dc.rights.access Open Access

Files in this item

This item appears in the following Collection(s)

Show simple item record Except where otherwise noted, this item's license is described as



My Account