Back to search results

PhD Studentship: Identifying and Understanding Complex Behaviour from Transactional Data

University of Nottingham - Nottingham University Business School

Qualification Type: PhD
Location: Nottingham
Funding for: UK Students, EU Students
Funding amount: £14,777 stipend plus UK/EU student tuition fees
Hours: Full Time
Placed On: 3rd December 2018
Closes: 23rd January 2019
Reference: SOC478

PhD studentship with Nottingham University Business School’s N/LAB.

We are looking for an enthusiastic PhD student who will undertake original multi-disciplinary research in the fields of data science/machine learning and consumer/human behaviour.

Supervised by N/LAB’s experts in data science/machine learning this PhD studentship provides an opportunity for people interested in developing novel statistical measures and methods to quantify, understand and leverage complex consumer behavioral regularity. Working with already acquired real world data (representing over 16 million regular customers) from N/LAB’s extensive group of collaborating industrial partners, the PhD provides an unparalleled opportunity to push forward research that will drive significant impact in predictive and descriptive analytical applications as well as contributing to knowledge regarding consumer/human behaviour.

Identifying and understanding regularity in consumer behaviour is a key component of business analytics. Examples range from customer/store segmentations (e.g. frequency) to automated predictive systems (e.g. automated email targeting) and management key performance indicators (e.g. repeat buying behaviour). However, regularity remains a tricky thing to measure. Traditional methods; such as average purchase frequency, repeat brand rate or statistics quantifying deviations from random interevent times are too coarse grained to adequately capture the true complexity of our ‘regularity’. This significantly reduces their utility as measures/features in end-use applications, such as customer segmentation/understanding (i.e. within marketing campaign development or management strategy planning) or customer behaviour prediction (i.e. churn prediction). Reasons for this failure can be roughly grouped into three categories:

  1. the presences of mediating factors (e.g. promotions, lifestyle, emotion, and life stage),
  2. incomplete observations and
  3. non-stationary behaviour. When these are not taken into account, regularity measures report high irregularity.

However, in reality the behaviour is regular once these factors are taken into account.

Recent advances in computational methods, and the systematic collection of consumer transactional data, now mean that the more complex measures of consumer behaviour regularity are within reach. The PhD will investigate:

  1. The development and extension of frameworks for quantifying consumer behaviour regularity within the context of mass transactional datasets and
  2. The development of novel statistical models.

Depending on the individual student, the developed methods are likely to deliver advances in one or more of the following application areas:

  • churn prediction with long prediction horizons,
  • consumer life stage identification and
  • prediction and/or understanding/predicting consumer variety seeking behaviour.

This funded PhD will commence from 1st October 2019. A yearly stipend at current RCUK rates (£14,777), plus UK/EU student tuition fees (£4,350), is available to the successful applicant.

Once your online application has been submitted, please email to indicate you would like to be considered for this scholarship.

We value your feedback on the quality of our adverts. If you have a comment to make about the overall quality of this advert, or its categorisation then please send us your feedback
Advert information

Type / Role:

Subject Area(s):


PhD tools
More PhDs from University of Nottingham

Show all PhDs for this organisation …

More PhDs like this
Join in and follow us

Browser Upgrade Recommended has been optimised for the latest browsers.

For the best user experience, we recommend viewing on one of the following:

Google Chrome Firefox Microsoft Edge