This project is a part of Udacity Machine Learning Engineer program. In this project I will apply unsupervised learning techniques on product spending data collected for customers of a wholesale distributor in Lisbon, Portugal to identify customer segments hidden in the data. I will first explore the data by selecting a small subset to sample and determine if any product categories highly correlate with one another. Afterwards, I will preprocess the data by scaling each product category and then identifying (and removing) unwanted outliers. With the good, clean customer spending data, I will apply PCA transformations to the data and implement clustering algorithms to segment the transformed customer data. Finally, I will compare the segmentation found with an additional labeling and consider ways this information could assist the wholesale distributor with future service changes.
This project is designed to give you a hands-on experience with unsupervised learning and work towards developing conclusions for a potential client on a real-world dataset. Many companies today collect vast amounts of data on customers and clientele, and have a strong desire to understand the meaningful relationships hidden in their customer base. Being equipped with this information can assist a company engineer future products and services that best satisfy the demands or needs of their customers.
- How to apply preprocessing techniques such as feature scaling and outlier detection.
- How to interpret data points that have been scaled, transformed, or reduced from PCA.
- How to analyze PCA dimensions and construct a new feature space.
- How to optimally cluster a set of data to find hidden patterns in a dataset.
- How to assess information given by cluster data and use it in a meaningful way.
A wholesale distributor recently tested a change to their delivery method for some customers, by moving from a morning delivery service five days a week to a cheaper evening delivery service three days a week. Initial testing did not discover any significant unsatisfactory results, so they implemented the cheaper option for all customers. Almost immediately, the distributor began getting complaints about the delivery service change and customers were canceling deliveries, losing the distributor more money than what was being saved. You've been hired by the wholesale distributor to find what types of customers they have to help them make better, more informed business decisions in the future. Your task is to use unsupervised learning techniques to see if any similarities exist between customers, and how to best segment customers into distinct categories.
This project uses the following software and Python libraries:
You will also need to have software installed to run and execute a Jupyter Notebook.
If you do not have Python installed yet, it is highly recommended that you install the Anaconda distribution of Python, which already has the above packages and more included. Make sure that you select the Python 2.7 installer and not the Python 3.x installer.
The customer segments data is included as a selection of 440 data points collected on data found from clients of a wholesale distributor in Lisbon, Portugal. More information can be found on the UCI Machine Learning Repository.
Note (m.u.) is shorthand for monetary units.
Features
Fresh: annual spending (m.u.) on fresh products (Continuous);Milk: annual spending (m.u.) on milk products (Continuous);Grocery: annual spending (m.u.) on grocery products (Continuous);Frozen: annual spending (m.u.) on frozen products (Continuous);Detergents_Paper: annual spending (m.u.) on detergents and paper products (Continuous);Delicatessen: annual spending (m.u.) on and delicatessen products (Continuous);Channel: {Hotel/Restaurant/Cafe - 1, Retail - 2} (Nominal)Region: {Lisbon - 1, Oporto - 2, or Other - 3} (Nominal)