FX Pair Price Movement Classification

Client Model

Front-office machine learning model to predict the direction of movement (up or down) and the price of currency pairs in the next minute.

Client Model Prediction

Predicted ClassDown
Model Confidence68.59%
R-Squared0.89
Significance0.0001

DC-MINT – AI EXPLAINABILITY

  • the direction of movement in time lag 30 was predicted as “Down”
  • The top three contributing factors towards the ‘Down’ prediction were – price at time lag 29, volatility at time lag 29 and trading volume at time lag 29
  • Provides the FX traders with the “reasons” behind specific predictions before acting on them
  • Provides the FX traders with the data that was used to “predict” prices and direction

DC-MINT INSIGHT