DESIGN AND IMPLEMENTATION OF PREDICTIVE MODELLING FOR IRIS,CHURN AND WEATHER DATA

INFORMATION SCIENCE AND ENGINEERING

The objective is to provide the data analysts or data science users to give much more powerful tool which would help them understand the patterns of data in better way. The tool would analyze the data and provide details such as outliers,imbalanced fields, missing values, performance of individual field and much more. This would help the data analysts to skip the manual process of cleaning the data before using the data to build a predictive model. Once the data is analyzed and cleansed,based on the business objective one can choose the outcome field that whose value should be predicted and then the tool would show the correlation between the different fields in the dataset. This would help the data analyst to remove the redundant fields from being used in the model building. Once the data analyst finalizes the fields to be used,a predictive model(Regression or DecisionTree) would be built from the tool automatically and would be available for download in PMML format (http://dmg.org/pmml/v4-3/GeneralStructure.html).One can download the model and use it to execute using any third party tools that are available in market like JPMML and other scoring engines. This would help the data analysts save time by removing the major manual task of data cleaning and model building.This tool also makes sure that it could be used by no-vice users as well.

This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.