Data Mining - Customer Retention – Telco Customer Churn Dataset

---

شرح پروژه

Tasks
Please complete the following tasks for this data mining project.
1. Data exploration (required): use proper summary statistics and visualization tools to
understand the distribution of individual variables and relationship between two or more
variables.
2. Data preprocessing (optional): detect and remove outliers, compute new variables from
existing variables if necessary.
3. Data and dimension reduction (optional): choose a subset of variables for model building
(dimension reduction), partition the records into homogeneous groups (using
unsupervised learning techniques such as cluster analysis) and build separate models
(data reduction).
4. Partition the data (required): use 10-fold cross validation to evaluate model performance.
Please refer to the description of k-fold cross validation in my slides on data mining
process (Chapter 2), and implement 10-fold cross validation using “for” loops.
5. Model building (required): build at least two classification models using only the
algorithms we have introduced in class, at least one of the built models must be
interpretable (in terms of model structure and parameters, which essentially reflects the
factors that contributed to customer’s decision to churn).
6. Model evaluation (required): evaluate performance of each model built. Specifically,
record and pool all training outcomes and validation outcomes from all folds separately,
and evaluate model performance using the following metrics and tools: accuracy,
sensitivity, specificity, precision, FDR, FOR, ROC chart, AUC of ROC curve, and Lift
chart. Compare the performance between training and validation and discuss whether
there is any overfitting issue or not.
7. Model deployment (required): choose one of the interpretable algorithms from step 5 and
build a model using the whole dataset. Interpret the model from both technical and

managerial perspectives.
 
Deliverables
1. A single report created using Microsoft Word format (Times New Roman font, 12 point
font size, 1.5 line space) that includes:
a. Cover page
b. A 1-page Executive Summary of the project
c. Detailed and organized documentation of description, key results (including
tables, charts, etc.), and discussions for each of the tasks you have completed
d. Convert your report to pdf for submission.
2. A single R file that includes code for all tasks you have completed, in their corresponding
order. 

این پروژه شامل 2 فایل مهم است، لطفا قبل از ارسال پیشنهاد حتما نسبت به بررسی این فایل اقدام فرمایید.

مهارت ها و تخصص های مورد نیاز

---

---