ABSTRACT
Financial companies came a long way to build massive data driven FinTech applications incorporating data mining capabilities via on-premises databases which now are in process of migrating to cloud computing environment. Moving to cloud, solved the computational power and space constraints. Finance industry is now turning their focus on to intelligent decision making systems using Machine Learning to predict the best possible investment strategies or trade ideas which can make prot to the clients. If machines can discover patterns in the trade, market data and suggest the best trade(s), or suggest portfolio balancing, suggest correction in the portfolios, trigger early warning on a set rules will certainly help investment managers. Our research and proposal of an Enterprise Architecture for Machine Learning with a case study of probability default modeling using Loan-To-Value (LTV) can help financial organizations to set up a their Machine Learning frameworks. Our proposal will address the key components of a framework by providing standards, work ows, auditable guidelines of models and support model at an enterprise level.
