Dear readers,
Since inception, we have been focusing our efforts on growing the research community on the Alphien platform. For the next 6-9 months, we will be embarking on the next phase for our platform with the launch of an Investor Portal catered for investment managers.
COMMUNITY
After concluding our last two academic projects, respectively covering volatility forecasting using a deep learning model and a traditional econometrics approach, and frequency analysis of price signals to detect market regimes, we reiterate our partnership with Télécom Paristech to tackle new research challenges. Students will work on forecasting gold prices using a panel of data science models, either supervised such as logistic regression and dense neural network, or unsupervised such as principal component analysis. The projects typically run for six months and are shared with the community to be used as a basis for further developments and fine-tuning.
In order to foster collaboration with academic partners and spread Alphien’s technology within universities, we are developing a framework to organise online challenges on the platform. This will allow professors to submit research subjects to their students on the platform, basically creating a virtual classroom for them. The objective is to incentivise professors and students to use Alphien as their main tool to do research, discuss new ideas and share results.
PLATFORM
As you already know, our platform is built on a complex IT infrastructure combining different web-based research tools hosted on our servers: AlphienStudio to write research scripts and functions, Notebooks to document research and display results, Qlib to load and enrich the financial function library and the Wiki to share, discuss and document technical aspects of quantitative research. For each tool, we have added new servers based in our Paris office, knocking two birds with one stone: first, scaling up capacity for our upcoming Alphathon; second, increasing speed for users based in Europe. The servers use a “two-way” synchronisation, meaning that changes made from one server location get reflected on the other server location, making our data storage processes and our service continuity extremely robust.
To further facilitate the sharing of ideas through code on our platform, we have enhanced the ‘fork’ functionality in our notebooks. Forking allows our community to build on what other users have already produced, and further improve on that existing work. It will provide our users with an added incentive to share the work they have done in order to get new insights from the community.
NEWSFLOOR
In this article on Medium, the author advocates that the future of trading belongs to Artificial Intelligence (AI). We agree with his views: “Computing already revolutionised financial trading once, it facilitated enormous numbers of calculations in a fraction of a second, and tracking markets that shift in light speed. Now AI trading systems are poised to foster a second wave of innovation, one that will be the most significant transformation in finance history.”
Unfortunately, the access to AI possibilities and benefits have traditionally been mostly restricted to giant corporations. But this reality is changing as AI becomes more mainstream.
And this is where Alphien gets in. Our platform gives investment firms access to the possibilities offered by artificial intelligence, machine learning, deep learning and more. We aim at becoming a key partner of asset managers in their AI transformation, connecting them with talented researchers and providing data scientists with state-of-the art data and programming tools. The AI revolution is underway in the financial sector and Alphien is surfing the wave!
