Improvement in Classification Algorithms through Model Stacking with the Consideration of their Correlation
Journal Title: International Journal of Advanced Computer Science & Applications - Year 2019, Vol 10, Issue 3
Abstract
In this research we analyzed the performance of some well-known classification algorithms in terms of their accuracy and proposed a methodology for model stacking on the basis of their correlation which improves the accuracy of these algorithms. We selected; Support Vector Machines (svm), Naïve Bayes (nb), k-Nearest Neighbors (knn), Generalized Linear Model (glm), Latent Discriminant Analysis (lda), gbm, Recursive Partitioning and Regression Trees (rpart), rda, Neural Networks (nnet) and Conditional Inference Trees (ctree) in our research and preformed analyses on three textual datasets of different sizes; Scopus 50,000 instances, IMDB Movie Reviews having 10,000 instances, Amazon Products Reviews having 1000 instances and Yelp dataset having 1000 instances. We used R-Studio for performing experiments. Results show that the performance of all algorithms increased at Meta level. Neural Networks achieved the best results with more than 25% improvement at Meta-Level and outperformed the other evaluated methods with an accuracy of 95.66%, and altogether our model gives far better results than individual algorithms’ performance.
Authors and Affiliations
Muhammad Azam, Dr. Tanvir Ahmed, Dr. M. Usman Hashmi, Rehan Ahmad, Abdul Manan, Fahad Sabah
Keywords
Related Articles
- EP ID EP499588
- DOI 10.14569/IJACSA.2019.0100360
- Views 108
- Downloads 0
Towards Security as a Service to Protect the Critical Resources of Mobile Computing Devices
Mobile computing is fast replacing the traditional computing paradigms by offering its users to exploit portable computations and context-aware communications. Despite the benefits of mobile computing, such as portabilit...
Ant Colony Optimization (ACO) based Routing Protocols for Wireless Sensor Networks (WSN): A Survey
Wireless Sensor Networks have several issues and challenges with regard to Energy Efficiency, Limited Computational capability, Routing Overhead, Packet Delivery and many more. Designing Energy Efficient Routing Protocol...
A Novel Position-based Sentiment Classification Algorithm for Facebook Comments
With the popularisation of social networks, people are now more at ease to share their thoughts, ideas, opinions and views about all kinds of topics on public platforms. Millions of users are connected each day on social...
Programming Technologies for the Development of Web-Based Platform for Digital Psychological Tools
The choice of the tools and programming technologies for information systems creation is relevant. For every projected system, it is necessary to define a number of criteria for development environment, used libraries an...
Smart Card Based Integrated Electronic Health Record System For Clinical Practice
Smart cards are used in information technologies as portable integrated devices with data storage and data processing capabilities. As in other fields, smart card use in health systems became popular due to their increas...