Generic Predictive Computational Modelling for Financial Data Use Case: Applying ontologies as financial data pre-processing tool
Speaker: Natalia Yerashenia, Alexander Bolotov
Abstract
This presentation introduces a comprehensive computational model for companies' bankruptcy prediction where ontology is utilised for data pre-processing (it defines the basic concepts of financial analysis, deals with data inconsistency, considering the semantic meaning and worthiness of data). We show how ontology can improve the efficiency of features selection to be fed as input for the machine learning engine. We will present and discuss a relevant use case where the proposed technique is applied to a financial dataset of a range of UK companies to assess their bankruptcy risk.
Slides
Bio
Natalia Yerashenia graduated with an MSc degree in Finance and Accounting from the University of Westminster (London, UK) in 2018. Currently, Natalia is the final-year Computer Science PhD Researcher at the University of Westminster. She is also teaching Mathematics for Computing Module for Computer Science students at the same university. Her interest lies in understanding how knowledge graphs, ontologies, and machine learning techniques can be applied in analysing financial and market data. Natalia's specialist topics are Fintech, Data Mining, Semantic Web / Ontologies, Graph Databases, and Neural Networks. She has contributed to several research projects including SMARTEST (https://smartestknowledge.org/) graph-based knowledge repository.
Dr. Alexander Bolotov is a Principal Lecturer and Software Systems Engineering research group leader. He holds a PhD in Logic from Lomonosov Moscow State University and a PhD in Computer Science from the Manchester Metropolitan University. Alexander has authored or co-authored more than eighty publications and edited several volumes of the conference proceedings. He is a chair of the UK Automated Reasoning Workshop, the leading UK forum in the area with 27 years of history. He has been involved in a variety of research projects including EC and Research UK funded projects. Alexander's research interests include mathematical modelling, formal specification and verification of reactive and distributed systems, and automated reasoning algorithms, Model Checking, System Modelling, Integration methodologies. His scholarly activities that are tightly linked to research, cover graph-based knowledge representation teaching of formal methods and digital repositories. Alexander has been recently led the development of the knowledge and learning repository SMARTEST (https://smartestknowledge.org/).