An offline workshop “Financial machine learning workshop: Using Mlfinlab, PortfolioLab and ArbitrageLab to create trading strategies”
On September 2, an offline workshop “Financial machine learning workshop: Using Mlfinlab, PortfolioLab and ArbitrageLab to create trading strategies’ ‘ will be held with Oleksandr Proskurin.
Speaker: Oleksandr Proskurin, Co-Founder and CIO at Principia Invest, Co-Founder at Hudson and Thames Quantitative Research.
Agenda:
Mlfinlab:
– Data Structures: compressing trades into bars (time, dollar, volume), Futures Roll
– labelling: triple barrier and trend scanning labelling. Meta-labelling and trend-following example on example of Bitcoin
– Feature importance analysis: MDI, MDA, Shapley feature importance on example of Trend-Following strategy on Bitcoin
Portfolio lab:
– Constructing investment portfolios using ML techniques and portfolio lab: HRP, HERC, NCO algorithms
– Denoising/Detoning correlation matrices
– Using Mlfinlab to get key strategy statistics: Sharpe, Probabilistic Sharpe, Max Drawdown, etc
