I am a quantitative model developer with a strong focus on Finance. Personlly I am fascinated by the advances that have been made in the recent years by Deep Learning models. Guided by my personal interests I compiled a collection of Deep Learning models into this
portfolio. The long term goal of this portfolio is to build a sense of how Deep Learning models may be transferred to applications in Financial research
as well as practical questions in Finance.
Reinforcement Learning is one of my favourite applications of Deep Learning. A famous text book example ist the cart pole problem, where the learner (i.e. the model) observes the evironment consisting of a cart that balances as pole by moving sideways. The learner applies horizontal force ranging form -1 to 1 in order to prevent the pole from falling over for 200 consecutive periods. The full implementation can be found in my projects below. The animation below shows the trained model sucessfully solving the cart pole task after 50'000 training episodes.