ANALYSIS OF CLASSIFICATION ALGORITHMS

Abstract

In this, I have analyzed and compared different classification algorithms and run on dataset and compared efficiency, and using combination of this algorithm shown to improve efficiency. I have used and trained model with four different classification algorithm and used those model to predict outcome of test data separately and then used combination of those model to improve production efficiency of test data prediction. With the advent of technology and machine learning development, there are various algorithms developed for model to be prepared which can take decision and predict something based on its learning. Many of research and analysis are based on these classification algorithms. It plays important role in machine learning. It helps to create model which can predict outcome based on its past learning. So if a model has multiple algorithms as a factor to decide or predict outcome it will be having good efficiency in order to predict right. As various classification algorithms exists so in order to pick right or suggest using combination of these algorithms in order to improve prediction efficiency. There are always a tradeoff between efficiency and execution time, here in this we have majorly focus on improving efficiency by using existing algorithms for a single model which consist of different models with different algorithms. Here in this I have majorly used skit-learn[1] libraries interface for various classification algorithms. I have used them in order to check and show efficiency of predicted output. By efficiency I mean with respect to correct output prediction. I have used various trained model for one single model in order to improve efficiency.