An automated cryptocurrency trading system based on the detection of unusual price movements with a Time-Series Clustering-Based approach

– Outlier detection significantly increases return on investment for trading strategies.
– Trading system is more profitable during highly volatile periods.

– Proposal of an end-to-end ML-based trading system with outlier detection module.
– Increased return on investment for machine learning-based trading strategies.

The paper proposes an ML-based trading system for cryptocurrency trading that includes a time series outlier detection module to detect unusual price formations and improve the profitability of trading strategies.

– Unusual price movements can negatively affect the functionality of technical signals.
– The generalization of the model may deteriorate due to unusual price formations.

– Time-Series Clustering-Based approach
– Classification algorithms for price direction prediction

– The proposed system increases return on investment for machine learning-based trading strategies.
– The trading system becomes more profitable during highly volatile periods.

– Outlier detection significantly increases return on investment for trading strategies.
– Trading system is more profitable during highly volatile periods compared to baseline model.

– ML-based trading system with outlier detection improves profitability
– Unusual price movements negatively affect technical signals

– The paper proposes an ML-based trading system for cryptocurrency markets.
– It includes a time series outlier detection module to improve profitability.