DATA SCIENCE PRACTICE PROJECT

PREDICT NEW USERS SUBSCRIPTIONS AND CUSTOMERS SUBSCRIPTIONS RENEWALS

INSTRUCTOR
Ardeshir Damavandi
Ardeshir is a Data Scientist based in Japan.
The last five years he has been working at Cognizant developing custom data models and algorithms to increase and optimize customer experience and  revenue generation.
He holds a Master of Computer Science.

PROJECT DESCRIPTION
This is a real life data science project. You are given a customer information dataset and you are requested  to classify clients based on the years of subscription and build a model to forecast new client’s subscriptions.
The project is divided in three main parts:
1) Data Cleansing
Data wrangling is a big part of any data science job. The data cleaning stage is done  through a series of assignments where you are asked to perform different functions and transformations in order to get the data frame ready to implement the Machine Learning algorithms.
2) Visualization
You are requested to plot several types of charts in order to explore data. 
3) Machine Learning
You will have to implement classification and predictive algorithms to predict subscriptions renewals   of current customers of a certain company. In the process, you will have to deal with categorical variables, poor accuracy rates, cross validation, data imbalance, and so on. 

DOWNLOAD / CONTENT
After purchase, you will receive an email with a protected  ZIP  and a password to access the content. If you are a  registered user, the download is always available on your account.
1) One PDF with the project description. 
2) One CSV, the dataset with customer information. 210947 rows and 12 fields.
3)  One jupyter Notebook with the project solved along with explanations about how the code works.

WHAT YOU WILL PRACTICE
– Libraries: you will have to work with the following libraries: pandas, numpy, matplotlib, plotly (free offline), re, datetime, sklearn.
– Python functions, loops (for), conditional statements (if/else) ,  dictionaries and  lists.
–  Pandas functions: groupby, dates, types, missing values, etc.
– Regular Expressions.
– Crosstabs.

REQUIREMENTS
You must be familiar with Jupyter Notebooks and Python data analysis libraries. If you do not have any of them,
it is recommended to install the Anaconda Package.
This project is aimed at Data Science students and anyone with knowledge of Python and Pandas. 
Moreover, the Machine Learning part entails the understanding of basic concepts about supervised learning.

CONTACT
If you need additional information, do not hesitate to contact us.