FlowHFT (Li et al. โ25, arXiv:2505.05784) merges dozens of expert HFT bots into one policy, topping the best expert in 90 % of simulated regimes.
๐น RL = tune your bot
๐น IRL = unmask their incentives
๐น IL = clone their strategy from data
@ammaar@Replit Google grounding is so expensive at $35 by 1000 quaries, and perplexity is around $5. Do you think there is a way to get around that price?
Excited to launch the first Crypto forecasting competition with Kaggle, G-Research, and Cambridge Spark. We have developed a very robust data set, and are looking forward to all the innovative solutions. #kaggle#cryptotrading#machinelearning https://t.co/Se2u7Hd6UM
[#PODCAST] The U.S. is mobilizing its friends & allies to rally against China. @ics_delhi analyst @vedavn joins us to discuss India's role in this burgeoning rivalry and where Africa fits in a rapidly changing geopolitical landscape.
Listen here: https://t.co/puNTHIROQQ
@TheDumbInvestor I agree anything robert touches turns into gold, to be frank I haven't added to that list in ages, its a shame, I now think these type of lists hold the field back unless they are regularly maintained and dated (send me a pull-request if you can)
Two for two today with our lecture on using machine learning to find viable pairs to trade:
https://t.co/q3uke0F2Ig
Based on the work of Sarmento, S.M. and Horta, N., 2020. Enhancing a Pairs Trading strategy with the application of Machine Learning.
By: @debrincat_aaron
@StuartReid1929 Have since pivoted a bit, but have sent you an advanced copy of what we are working on LNKD (I have some reservation still) let me know if you see anything untoward.
Women in AI and Finance workshop sign up here for free, it is a great line up and the first of its kind. It is virtual but registration is required. Details at
https://t.co/CYc4NsY1Kp