FOLLOW

FOLLOW

SHARE

Speaker Snapshot: "Banks Must Allow For “Challenge Mechanisms.”"

We speak to Nikhil Aggarwal, FinTech Entrepreneur in Residence at iValley Innovation Center

1Dec

Ahead of his presentation at the Machine Learning Innovation Summit in New York on December 11 & 12, we spoke to Nikhil Aggarwal, FinTech Entrepreneur in Residence at iValley Innovation Center.

Why do you think we have seen machine learning use increase so dramatically in the past 3 years?

Machine learning (ML) is not a new issue. In fact, the origins of machine learning techniques go back to the 1950s when the first learning machine, first neural network machine and perceptron algorithm were invented.

In recent years, the increase in adoption of machine learning can primarily be attributed to i) application of ML to a wide range of practical problems and use cases, ii) explosion in data (volume, variety, velocity and veracity) leading to significant improvements in model accuracy and predictability, and iii) increased conversation and action by non-technical stakeholders to experiment. For instance, in Banking, we are now using machine learning approaches to discover incremental revenue opportunities, enhance the customer experience, better manage the identification and mitigation of varied risks; and to automate simple processes.

How do you think organizations could be utilizing machine learning better?

Defining a business case at a more granular level, working through the data construct and building out meaningful hypotheses will better position an organization to further adopt machine learning. Reviewing past modeling efforts (both successes and failures) also plays a key role in enabling organizations to better utilize machine learning.

Organizations also need to “pair up” domain subject matter experts (SMEs) with data scientists so that both groups can better contribute to defining the inherent business challenge and arrive at a viable solution. For instance, you must have historical “goods” and “bads” on a supervised learning problem such as predicting which customer sub-segment will have a higher fraud loss metric. In an unsupervised problem, a machine may spew out granular clusters based on pattern recognition and discovery. This classification may have to be regrouped before any business decisions can be made.

What are the biggest challenges currently facing the further spread of machine learning?

Let us take the example of an ML Team which claims that they are using deep learning and artificial intelligence to reduce false positives by over 60%. Their pitch while brilliant is overly technical and complex. The core challenge here is that the ML Team has not bridged the “gap” between risk management/business leadership and technologists. Often, the value proposition focuses on a predefined solution utilizing a particular machine learning technique which may not capture a broader set of evolving risk nuances. As a result, a proposed solution may check the technology boxes, but will not address all the underlying regulatory, compliance and operational risks.

To increase the adoption of machine learning, practitioners must explain their solutions and answer the fundamental questions on how core issues are being addressed. The solution must be implementable.

Do you think that machine learning regulation is currently fit for purpose?

In the financial services industry, most institutions have large model validation, testing and audit teams to ensure that ML models comply with existing regulation.

Banks must allow for “challenge mechanisms.” For example, if a “fit for purpose” model suggests that product X is the “best one” for customer Y and if this does not make business sense; override decisions must be permitted.

What can the audience expect to take away from your presentation in New York?

I will cover an overview of the applications of machine learning in the identification and mitigation of money laundering and financial crime. Increasingly complex regulations coupled with heightened scrutiny from regulators have resulted in more banks and financial institutions building out broad spectrum anti-money laundering programs. Although efforts continue to focus on hygiene factors, such as building out data lakes and reporting routines, there is now a carefully considered interest in experimenting with advanced analytics, machine learning and artificial intelligence techniques.

You can check out Nikhil's presentation at the Machine Learning Innovation Summit in New York on December 11 & 12.

Comments

comments powered byDisqus
Looking out small

Read next:

Big Data Top Trends 2018

i