Internship research - Adaptive online learning from expert advice

ABOUT CFM

Founded in 1991, we are a global quantitative and systematic asset management firm applying a scientific approach to finance to develop alternative investment strategies that create value for our clients.
We value innovation, dedication, collaboration, and the ability to make an impact. Together, we create a stimulating environment for talented and passionate experts in research, technology, and business to explore new ideas and challenge existing assumptions.

Overview and goals

The project is to review the state of the art of learning allocation algorithm in well controlled limit cases spanning stationary high-/ and low- S/N ratios, and non-stationary high-/ and low- S/N ratios. The scope of the algos spans mean-variance optimization and its regularized versions, simple on-line learning (e.g. FTL, Fixed Share) and less simple time-adaptive online experts allocation algorithms (e.g. ensembles or AdaNormalHedge) with and without performance predictor.

Once review and testing of some known models is complete, the research will move to new directions. Possibilities are the redefinition of the loss-function to include risk and/or trading costs, or the estimation of alternative “solutions” to previous ones.

The final goal is to understand which algorithm to use in each context of stationarity and/or S/N ratio up to coding such system in a meta-learner.

The research will be done only on synthetic data; with possible application to real data if synthetic results are convincing enough.

Profile

Student on a masters degree course like applied mathematics, computer science, theoretical physics

Descent level of Python required.

Finance background not necessary.

References

• [arXiv: 1909.05207] Introduction to Online Convex Optimization
• [arXiv: 1502.05934] Achieving All with No Parameters: AdaNormalHedge
• [arXiv: 1103.0949] Adapting to Non-stationarity with Growing Expert Ensembles
• [JMLR: D. Adamskiy et al, 2016] A closer look at adaptive regret
• [arXiv:1208.3728] Online Learning with Predictable Sequences

Duration

4 to 6 months

EQUAL OPPORTUNITIES STATEMENT

We are continuously striving to be an equal opportunity employer and we prohibit any discrimination based on sex, disability, origin, sexual orientation, gender identity, age, race, or religion. We believe that our diversity, breadth of experience, and multiple points of view are among the leading factors in our success.
CFM is a signatory of the Women Empowerment Principles.
 

FOLLOW US

Follow us on Twitter or LinkedIn or visit our website to find out more about CFM.