Telemarketting success strategy from machine learning perspective

Machine learning (ML) provides us the flair to achieve knowledge in different areas from large data pools that we as human beings may not study. So, the ML is a process in which we train computers to recognize patterns in a dataset, such that they learn that significant information and use them to achieve new things that they haven't encountered before. The main hypothesis covered in this thesis is as follows: feature selection in supervised classification tasks can be performed on the basis of correlation between features and such a selection process can be beneficial for various machine learning algorithms. On this purpose, we turned our attention on investigating whether successful predictions can be achieved for a company's decision-support mechanism by analyzing a selected telemarketing dataset in a machine learning algorithm. In conclusion, this thesis reveals that the feature selection in a supervised machine learning analysis can be performed on the basis of correlation among the corresponding attributes.