Analyse an online retail dataset for customers segmentation. Link to the data: https://archive.ics.uci.edu/ml/datasets/online+retail
If the file cannot be renderred in github (sometimes it happens!), please use this link instead: https://nbviewer.jupyter.org/github/linh279/customers_segment/blob/master/Clustering_Retail_Data.ipynb
We want to use this data to study the problem of customers segmentation, which is a common task that businesses, especially retailers, require in order to undertake marketing activities, pricing plans, customer services and other business strategies.
We are going to do the followings:
- Data preparation: the dataset has different issues in terms of duplicates, problematic entries, missing entries and outliers that need to be dealt with
- Feature creation: derive extra features from the data (frequency and equally-sized price bins)
- Exploratory analysis
- Data transformation: transform the data to be suitable for clustering algorithms and dimension reduction.
- Dimension reduction for feature selection
- Clustering using different techniques
- Evaluation of clustering results