Hybrid Rule Based Technique for Geo Spatial Data Processing and Data Analytics using Machine Learning Techniques

Abstract

The Data Analysis and Prediction of results are the two major steps involved in Data Science
at high level, while making the decisions with respect to any domain or business. While
coming to Geo Spatial data these two steps are much more important as they have to be very
clear amongst the existing huge data. In order to achieve the accuracy in Data Analysis and
Prediction with respect to Geospatial data, the probabilistic models are not so accurate as
most of the times the data is collected from various resources and it needs to be segregated
based on set of rules and the final results should be ready to get accurate prediction of the
outputs which are expected at business level. In this proposed research a Hybrid Rule Based
Technique is introduced by considering all the obstacles the researches are going to face
whiles using the cleaned geospatial data for further classification to implement in real time
system. The technique followed in this research will create a rule based technique along with
probabilistic methods in order to improve the accuracy in classifying the huge data. The
hybrid rule based technique majorly concentrates on the improving the quality of the output
data which is going to be generated as per the need of the researchers by which the throughput
of the final system will be improved. And it makes the top level management to take accurate
decisions and also to achieve the expected targets in the business. These hybrid rules will be
again stored in the form of a block of codes and it will be chosen based on the packages like
Scipy which is used in Machine Learning techniques.