AutoTrader Industry Problem Solving Event
Alan Turing Building
22 February 1-3pm

The Auto Trader Day is an industry problem solving event for undergraduate and postgraduate students from The University of Manchester. Auto Trader is a company which lists cars for sale, with over 437,000 cars listed and 48 million visits each month, quantitative methods are essential to managing their platform. The challenge is to apply your mathematical (and possibly programming) skills to solve some real-world problems that they face. You will gain insight into how your maths is likely to be applied once your studies are over.

Fill the registration form below to let us know you will be there. Note that you need to be a member of the Manchester SIAM-IMA Student Chapter to attend this event. If you have not joined the Chapter yet, you can register here.

Problem outlines

1. Curve fitting for car valuations
We have data on sold prices of different kinds of car. What is the best valuation model we can formulate by fitting different curves, in different ways, to the data?
2. Clustering user search behaviour
Users of the Auto Trader website find cars by entering search filters, for example make, model, age and price. Are some combinations more common? How can we explore the structure of the search space, and show if there are distinct clusters in the data?


13:00 - 13:20Introduction to Auto Trader and problems
13:20 - 13:40Regression
13:40 - 14:00Clustering
14:00 - 15:00Give it a go at the computer cluster
15:00 - Discussion and prizes


Dietary requirements



Very good introductions to Regression analysis and K-means clustering can be found on wikipedia.

Click here for the slides on linear Regression. The corresponding script can be downloaded here or viewed as an iPython notebook on Jupyter here (alternatively download the notebook here).


The data sets for the event can be found here:

To load the data into matlab use:

    >> A1 = csvread('Ford_Focus_prices.csv' , 1 , 2);
    >> A2 = csvread('Viewed_ads.csv' , 1 , 1);

Organising Committee

Jonathan Deakin
Massimiliano Fasi
Matthew Gwynne
Georgia Lynott
Mante Zemaityte