A Study on Analysing the impact of Feature Selection on Predictive Machine Learning Algorithms

Abstract

In recent times, one of the most used scenarios in many industry domains is enhancing the bids or tenders made by suppliers. In this paper, we will be analyzing one such use case for studying the effects of mixed feature selection to optimize the Learning model. The use case is to target and build a predictive clustering model in such a way that the scheduler receives the suggestions based on the most optimal options. There are few feature selection, enhancement, and scaling methodologies which this paper aims to explore with real-time data. Based on the analysis, the most important feature derived would be used to predict the optimal suggestion. The results will then be compared to understand the shortfalls and strong points of this new approach based on the accuracy of prediction. A clustering model will not just help reduce the hours of manual effort put into selecting the right source but will also provide an authentic and optimal option for a scheduler's consideration.