Conceptual study-A review on various machine learning algorithms of datamining

EasyChair Preprint no. 3103

Abstract

Rising concerns of database industry and resulting market needs for various methods to extract valuable knowledge from large data stores, thus data mining DM and KDD has emerged as a problem solving tool for analyzing data for the databases that are preexisting. This paper reviews on various machine learning algorithm used on various training data set used from UCI repository. Machine learning is categorized as supervised and unsupervised learning, accordingly supervised learning is obtained from various classified concepts i.e. classifier for new instance. Unsupervised learning concerns on various unclassified class. Predictive datamining is also called as supervised learning and descriptive datamining is unsupervised based on various association rules. Machine learning and Datamining approach focuses on analysis of categorical, on-numeric data and on the interpretable data. Cross industry standard process for datamining CRISP-DM a mining techniques involved for business solutions based on KDD. A review is done from various research papers on datamining tools and algorithms and its effect on supervised learning for fruitful decisions on data.

Keyphrases: Datamining(DM), Key words: UCI, knowledgediscoverydatabasesKDD, supervised, unsupervised

@Booklet{EasyChair:3103,
  author = {C. B. Lakshmi},
  title = {Conceptual study-A review on various machine learning algorithms of datamining},
  howpublished = {EasyChair Preprint no. 3103},

  year = {EasyChair, 2020}}