期刊名称：International Journal of Innovative Research in Science, Engineering and Technology
印刷版ISSN：2347-6710
电子版ISSN：2319-8753
出版年度：2018
卷号：7
期号：8
页码：8645-8652
DOI：10.15680/IJIRSET.2018.0708059
出版社：S&S Publications
摘要：Sequence classification is an essential undertaking in information mining. We address the issue of
sequence classification utilizing rules made out of intriguing examples found in a dataset of named sequences and
going with class marks. We measure the intriguing quality of an example in a given class of sequences by joining the
attachment and the help of the example. We utilize the found examples to produce sure classification guidelines, and
present two distinctive methods for building a classifier. The principal classifier depends on an enhanced rendition of
the current technique for classification dependent on affiliation rules, while the second positions the principles by first
measuring their esteem particular to the new information question. Experimental results demonstrate that our
administer based classifiers beat existing equivalent classifiers regarding exactness and soundness. Also, we test
various example include based models that utilization various types of examples as highlights to speak to each
sequence as an element vector. We at that point apply an assortment of machine learning calculations for sequence
classification, experimentally showing that the examples we find speak to the sequences well, and demonstrate
compelling for the classification undertaking.
关键词：Sequence Classification; Itemset; Classification based on associations; Sequence Classification
dependent on Interesting Patterns; dataset;
