6sense_model_predict

Full project: https://github.com/YingluDeng/6SenseCallPrediction

Introduction

This notebook is for training the 6sense data through different models (such as Random Froest, KNN, XGBoost and so on) and find the best fit model through several variables including call disposition, best day of week to call and best hour to call, predicting the success or not if a customer would pick up phone calls so that the company can get engagement with prospective customers.

The Data:

• calls.csv: a timeline of outgoing sales calls and the disposition of those calls
• events.csv: any activities that we have on record taking place before the phone calls were made
• companies.csv: the industry and employee count of the companies
• people.csv: the people who were called, along with their job level and function and the ID of the company they work for
• opportunities.csv: the date an opportunity was generated for a contact

Tableau Visualization (Merge Table):

Tableau Visualization For Each 6sense Csv: