Synthetic Data & Financial Crime: A Missing Cog In The Machine? (Webinar)
You are cordially invited to a FinCrime Dynamics (formerly EalaX) presentation hosted by Agent Based Modelling London with keynote speaker Daniel Turner-Szymkiewicz, outlining how we can find value in synthetic data with machine learning and its many applications towards the combating of financial crime.
Register to get access to our webinar in Synthetic Data & Financial Crime: A missing cog in the machine?
ABM London Meetup Registration
Speaker Bio:
Graduate of BSc Pharmacology and MSc Neuroscience with an interest in the analysis and refinement of synthetic data through machine learning in the field of financial crime. Currently working as a Data Scientist for FinCrime Dynamics (formerly EalaX) in the application of machine learning models on synthetic data in fraud detection. His other interests and pursuits are; explainability in machine learning models to address bias and fairness, application and viability of synthetic data to create robust clinical trial models, and neuro-pharmacological and immuno-pharmacological targeting of brain haemodynamics and metabolism in the treatment of neurodegenerative disease.
Talk Abstract:
Through the covid-19 pandemic, we have seen a societal shift, whereby more and more people are now being pushed into a digital ecosystem. With such a drastic shift in consumer activity, we also see an evolution of fraud occur that takes advantage of vulnerabilities that appear in the wake of such dramatic changes.
Financial fraud requires extensive knowledge of protocols and systems to access accounts and transfer services, thus banks have been compelled to increase investment in security and fraud protection. One of the greatest challenges financial institutions face today is to rapidly adapt their control systems in wake of these changes. These institutions tune their control systems according to applicable regulations, which carries two clear objectives: detect and prevent as much criminal activity as possible (through increasing true positives), and reducing the number of innocent people wrongfully accused (through the reduction of false positives).
Machine learning (ML) fraud detection algorithms, including recent advances in deep learning (DL), can play a more effective role in finding the hidden correlations between user behaviour and fraudulent actions and can evolve with minimum input from analysts. However, the establishment of an ML governance framework is challenging, and the adoption of technical standards to create better practices is still an ongoing process in the industry.
This talk explores The value that can be derived from synthetic data for financial crime, and how its use can be validated for the purposes of machine learning. Featured, is a compelling story on the inner workings of money laundering operations; and how synthetic data can be used to fight back against sophisticated crimes such as these. Moreover, it examines the ways in which synthetic data can be benchmarked for improving financial crime detection.