education

Education


Round Corner
Department of Computer and Information Science

Oppgaveforslag

Bankrupcy prediction (AI Lab pitch)

Background: This project/master thesis proposal is a collaboration between the  Norwegian Open AI-Lab and BDO.BDO provides a range of services within the areas of Audit, Accounting, Consulting and Tax/Legal. BDO Norway employs more than 1450 people and has more than 70 offices throughout the country. Our clients include major, global companies to small and medium-sized enterprises Internationally, in most areas of the private and public sectors. BDO is
present in 162 countries and employs more than 74 000 people.

The Task

BDO has access to a lot of accounting data and works systematically to ensure we are able to offer data analysis across our service lines. To provide further value based on the data we have access to, we will attempt to apply AI on the data. 

The specific task we want to collaborate with NTNU on is bankruptcy prediction. Our hypothesis is that it is possible to provide predictive analysis, using machine learning (ML) to create a model based on a set of publicly available accounting data aggregated at the year-level. The model would then be applied to more granular accounting data aggregated at the same level. A subtask will be to determine how and to what extent the model is applicable to such aggregated, granular data. Furthermore, the value of the model will increase if it can explain why it predicts as it does.


During the spring of 2019, BDO collaborated with a group of international master students at Norway’s top economics university NHH. Based on the collaboration, we will have a report detailing domain knowledge about data and bankruptcy prediction. This will be available to the NTNU students and we hope the domain knowledge on economic theory of accounting data and bankruptcy prediction from NHH together with technical
knowledge from NTNU will prove a strong combination and will ease the work with feature engineering.

The Data
BDO collect and store data in a data warehouse. The data that will be used to create the model will be data from all companies in Norway. The data will contain companies still in business and companies that have gone bankrupt. The dataset will contain financial records for several years back in addition to other data about the company such as shareholders, share capital, announcements, line of business etc. 

 

The data resides in our data warehouse and can either be accessed directly or be extracted to a file. The dataset is not publicly available, and access/extraction is subject to approval from our Risk & Compliance department. A Non-Disclosure Agreement (NDA) will need to be signed bythose who get access to the dataset(s).

 

The Practicalities
The work will be organized from the department of Strategy and business development (SFU), which is BDO's innovation spearhead. We would prefer if the students in the project are given responsibility both for leading the project and for producing the content.


Resources from Data & Analytics and Innovation in SFU will assist with the work. However, please note that machine learning is a new field for the department, and as such there is limited access to technical resources from BDO.


Although most of SFU is located in Oslo, students working on this project can use our offices in Trondheim (Lerkendal) and get access to computers and data to complete the task.

 

Contact at BDO:

KARL KLEIVE
Senior Manager
Strategy and business development | Data & Analytics
BDO AS
(+47) 91 86 86 15
karl.kleive@bdo.no

 

Other information:

This is a project in collaboration with an external partner. If you choose this project, then I will serve as the responsible from NTNUs side, but the actual work will also be in tight collaboration with personell from the external partner as listed above.

If you consider picking a project with me as the supervisor, then please look at www.idi.ntnu.no/~helgel/teaching/proposals/.

 

 

Faglærer

Helge Langseth Helge Langseth
Professor
310 IT-bygget
735 96488 
 
NTNU logo