Objectives
The objective of the practice module is two-fold:
• Firstly, it exposes participants to real world financial services problems so that they can learn to practice the skills they have gained during the component courses in a holistic manner.
• Secondly, it enables participants to demonstrate their proficiency across all the skills that they have learned in the course modules and hence obtain a grade at the Graduate Certificate Level.
Components
There are two parts in the Practice Module.
- Practice Project:
Participants will need to undertake one or more projects to gain practical experience and demonstrate their understanding and mastery of the skills taught in the four component courses. The practice project will require each participant to expend an estimated 10 man days of effort. These days are not expected to be contiguous and may stretch over many weeks and months. These projects may be conducted by individual participants or in teams depending on the nature of the project requirements. Participants are expected to understand business requirements for DS/AI in financial services projects, identify multiple data sources, build sophisticated analytics model and apply key data analytic techniques to find out the insights and solutions.
- Examination:
Each participant is required to sit for an examination on a stipulated date and time.
- The overall grade for the participant will be based on the Practice Project and Examination.
Typical examples of projects to be undertaken
One of the potential project can be in credit scoring using traditional and alternative data. Financial institutions have long relied on traditional data sources for credit scoring to assess creditworthiness of individuals and companies. In the recent times, the proliferation of data sources such as online shopping patterns, telco usage, social media activity, geolocation data, etc., presents a trove of new data sources that provide alternative richer insights into consumer behaviour and financial risk. These alternative data can be used to complement the traditional data sources and for credit scoring decisions, for making micro-loans accessible to the underserved communities or the unbanked sector. If a black-box technique is used to develop the model, explainable AI can be used to make the solution regulator friendly.
Financial fraud detection has always been a key issue in the financial services sector where great emphasis and effort has been put in to counter such fraud when detected. Such financial fraudulent activities could take the form of credit card fraud, insurance claim fraud, trading fraud, online banking fraud, etc. These can be done via a variety of methods and such activities may be hard to detect. With the increase in sophistication of the fraud techniques over the years, advanced innovative techniques in AI/DS has been used more pervasively in recent years to counter such activities.
Traditionally, insurers have been using information from client filled application forms and medical examinations for insurance underwriting. Typically, underwriters will evaluate the age, gender, BMI, answered medical history, family medical history, etc. to access whether to approval the client’s application or to calculate the insurance premium required. Additionally, a visit to the medical doctor and blood tests may also be required, and this process takes time. Alternative data such as biometric wearables, electronic health records, etc. may provide significant advantages to complement this underwriting process. These alternative data together with traditional data will give a more holistic view of the applicant and hence a more accurate underwriting. With the addition of alternative data, the overall data volume to be processed would be larger, and DS/AI model driven insurance underwriting will allow for a more efficient and quicker process.
Deliverables and success criteria:
- select the appropriate DS/AI analytics techniques for the identified financial services problem statement
- implement DS/AI analytics techniques taught within this certificate
- adopt explainable and responsible AI practices
- interpret analytical results and demonstrate the insights from the results