Developing a Statbot for Predictive Analytics


  • Suprateek Halsana
  • Vedant Singh
  • Sumit Upadhyay
  • Zubair Ahmad


Automatic Statbot for predictive analytics is the future of the data analytics world where we would have trillions of unprocessed data which needs to be structured, preprocessed, and to be converted into a form which could be used up by the sophisticated users like data scientists, data analysts, business analysts, etc. The Statbot interacts with its user/client in a friendly manner and asks for the dataset to be used for the predictive analytics. Thereafter, it requests the user/client to tell what is to be predicted. Being automated, it generates its statistical computations and within a few seconds preprocesses the dataset. The same work done by an analyst would have taken couple of hours to several days. Hence, it is undoubtedly superfast and efficient in its work. After the data being preprocessed, it intelligently finds out the best possible machine learning algorithm for the predictive analysis. It even keeps an option to perform a specific machine learning algorithm of our choice. In addition to this, it even makes our work verifiable by presenting us the report of scores and errors to be checked if required. The visualization an analyst dreams of to make the clients satisfied is now easily obtained as the Statbot provides an excellent visualization to focus on all the statistical information a client holds interest with ease and good interpretation. The main objective of this research is to showcase the various modules and algorithms designed and utilized to make the overall backend of this project. It also aims toward discussing the various outcomes and results we encountered during the testing of the statbot.