Soham

Currently a Computer Science student doing an Meng in Computer Science. I am looking to specialise in machine learning and financial computation.

I enjoy applying maths to the world of CompSci and so this got me interested in machine learning and statistical learning. Finance is also a passion of mine so I try to find ways to combine my interests. This resulted in me undertaking my current project, a transformer-based model that has been modified to be used for time series data, the Temporal Fusion Transformer. I want to do further research and exploration into the Fourier series and the subsequent Fourier analysis. I am also looking into the role of computers and algorithms in the world of finance. I have worked in Python, Java, HTML, Javascript, and Latex.

Last Summer I interned with Mako Trading where I gained a better understanding of the financial landscape and have gained a greater appreciation for the world of optionality. Inspired, I am looking to explore algorithmic trading, particularly the application of machine learning in this space. My next few projects will involve an exploration into Numerical Methods, Deep Hedging, and an implementation of the TFT in this space.

During Spring, I was offered Spring Insight Weeks at Citi, Barings, Macquarie, and JP Morgan, completing the first three. These gave me a great insight into the world of investment banking and the uses of quantitative finance within them! Last Summer I interned at Maker's Lab Tech, Mahindra as a Quantum Computing Researcher. This included exploring the uses of Quantum GANs, Quantum Circuit Born Machines and several numerical methods using Quantum techniques. One of the main projects undertaken was an exploration into synthetically generating market data by learning a given distribution.

I have finished researching Deep Hedging using Quantum Techniques through my third-year dissertation, and the resulting paper can be found on my page.

This Summer, I interned at Macquarie in the Commodities and Global Markets division as a data science intern. Here I was responsible for using deep learning to forecast renewables demand and price North American power markets(ERCOT).

During my Master's year, I intend to research Quantum Monte-Carlo for calibrating options pricing models in energy markets; I hypothesise a quadratic speed-up in parameter estimation over classical methods.