Question 1

How does the platform use machine learning?

Accepted Answer

We run a 20-model ensemble across 173+ assets — LSTM networks, XGBoost, CatBoost, Random Forest, and Transformer architectures, all trained on our own historical data. The ensemble generates directional signals by aggregating predictions from models that each specialise in different market regimes. Think of it as a quantitative research desk, automated.

Question 2

Can I actually learn ML for trading here?

Accepted Answer

We have a full education hub — 25+ structured lessons covering supervised and unsupervised learning, neural networks, time series forecasting (ARIMA through to Transformers), reinforcement learning (DQN, policy gradients), risk management (VaR, drawdown analysis), and portfolio optimisation (Markowitz, risk parity). All taught with financial data and real examples, not toy datasets.