Event Agenda

Agenda

Machine Learning for Risk Managers: Agenda

**************************************************
Live virtual course | Below agenda timing is in BST

Respective time in HKT:

Start: 8am BST | 3pm HKT
Break: 9am BST | 4pm HKT
End: 10.15am BST | 5.15pm HKT

**************************************************

08:0009:00

Applications of machine learning in risk management

08:00 - 09:00

  • What is machine learning?

  • Brief overview of the four main areas of machine learning

  • Front line uses of machine learning in risk prevention and detection

  • Data and model risk management with AI

  • AI in the operational risk function

  • Discussion: potential beneficial use-cases for AI in your firm 

09:0009:15

Break

15:00 - 15:15

09:1510:15

Challenges in AI implementation

09:15 - 10:15

  • Skill deficit

  • Finding beneficial use cases

  • Ensuring employee support

  • Managing AI risks 

  • Joint effort requirements of AI

  • Discussion:  main challenges to applying machine learning tools in your firm

08:0009:00

Case study 1: text analysis & Event Categorization Model

08:00 - 09:00

  • The virtue of ML: reading text

  • Using textual data to build categorisation models

  • Determining model inputs and measuring model performance  

  • Handling imbalanced target data 

  • Ensuring model alignment with firm risk appetite

  • Discussion: how could you adapt a general event categorization model to your firm

09:0009:15

Break

15:00 - 15:15

09:1510:15

Case study 2: predicting program and project performance

09:15 - 10:15

  • Improving your predictions with Machine Learning 

  • Overcoming challenges: using small datasets for machine learning

  • Combining data types for input into the model

  • Designing a prediction model to improve performance

  • Discussion: define potential long and short-term predictions a model could make.

08:0009:00

Machine learning model risk management

08:00 - 09:00

  • Overview: What is model risk management?

  • Understanding the machine learning life-cycle

  • Potential data and model issues

  • Maintaining resilience when models fail 

  • Discussion: analyse potential biases in different hypothetical models and data 

09:0009:15

Break

15:00 - 15:15

09:1510:15

Managers’ role in mitigating machine learning risks

09:15 - 10:15

  • Establish robust data governance procedures

  • Assess data quality

  • Evaluating the risk posed by a model

  • Ensure model monitoring and validation align with risk appetite

  • Discussion: assess the potential risks posed by a hypothetical loan applicant approval model 

10:1510:15

End of course

16:15 - 16:16