|
MR-mRMR Feature Selection Approach with an Incremental Classifier Model in Big data
|
|
| Author:
|
BLESSY LINCY.S.S, SURESH KUMAR NAGARAJAN
|
| Abstract:
|
The selection of the features for the processing plays a vital role which also has a significant impact on the overall task
that has to be performed with the data. With the rapid technological growth the data handling approaches and
methods has evolved to meet the demands faced with the voluminous data. In this paper, a feature selection algorithm
is proposed to choose a subset of features that represent the entire dataset for the processing. It is evident that the
data quality can be improved on applying the feature selection. Further, an incremental model is applied on the data to
perform classification and to handle the new arriving instances of the data. The proposed approaches are
implemented along the Apache Spark framework and the results are analyzed to show the efficiency in terms of
various performance metrics. The experimental study made with the set of real life massive data streams shows the
effectiveness of the proposed approach and it is evident that we are able to provide the efficient nearest neighbor
method for the high-speed big data with streaming data instances.
|
| Keyword:
|
MR-mRMR algorithm, big data, Incremental model, Classification
|
| EOI:
|
-
|
| DOI:
|
-
|
| Download:
|
Request For Article
|
|
|