#我的策略演变 My trading strategy has undergone three paradigm iterations: Initially relying on moving average crossovers and RSI oversold signals, I fell into the overfitting trap, with an annualized return volatility of 45%; In the mid-term, I integrated fundamental cycle judgment with on-chain data to construct a multi-factor quantitative model, improving the Sharpe ratio to 1.8, but neglecting market sentiment resonance led to a liquidation during the 2024 black swan event; Currently, I am shifting to a dynamic adaptive framework, introducing LSTM to predict market sentiment curves, combined with the Kelly criterion for dynamic asset allocation, validated through cross-cycle stress testing, with an annualized return stabilizing at 27% and maximum drawdown controlled at 8%. The core shift is from mechanical rule compliance to dynamic balance of risk-reward ratio, fundamentally evolving from linear causality to a nonlinear complex system.