Advanced Machine Learning for Financial Services

About This Course

AI/ML models in Banking (along with Finance in general) industry are becoming increasingly popular due to the following reasons:

1) Hope to have sharper discrimination with AI/ML models so that more customers can be approved using scientific tools (e.g. credit score) while satisfying regulatory demand
2) Unregulated lending organizations are using AI/ML techniques (as they don’t need to make regulators happy) rampantly, using open source tools and hence adopting their approach (herd mentality & fear of losing). Traditionally banks have been using commercial tools (primarily SAS).
3) For last 25 years remuneration growth for banking sector has been distinctly higher compared to the other sectors and hence growth in cost of hiring (salary) has become unsustainable. However, after GFC in 2008 the workload has increased tremendously due to heightened regulations. So, banks need to manage more work with less people. So, automation needs to be adopted in as many functions as possible. Hence in addition to underwriting and account management, the areas like fraud detection, document classification, operations where AI/ML techniques are being explored for process improvement and reduction in manpower requirement.
4) In olden days, only structured data used to get processed, but with the emergence of Big Data technology customer management with social media data etc. are becoming increasingly popular.
5) Technological growth is making complex model implementation much simpler.


In terms of analytics adoption, many a financial firms are now heavily investing in this area. This course is a part of this initiative from NUS-ISS to support such initiatives.

This course is designed to meet the need of financial firms to understand advanced ML models (beyond basics) which they are either using now or plan to use in future. As an example, banks perform the role of a “checker” i.e. whatever model is used by risk management or other departments for underwriting, account management, fraud detection, document classification, news/information classification, operations etc. they need to validate that they are developed correctly and adhere to either regulatory requirement or comply appropriately to bank’s operations guideline. As the model developing units of a bank are venturing more and more into advanced ML models, understanding the mechanics behind them is becoming increasingly relevant. After attending NUS-ISS’s SB, PA & CA courses for the preliminary grounding, this course will elevate participants’ learning process further in terms of other advanced topics and for companies this will in-turn mean a task force capable of applying the relevant financial analytics techniques with suitable technical and contextual reasoning.