learned stacked-ensemble family blend
In plain terms
A machine-learning model studies how all our best strategies for a stock have behaved and learns the smartest way to combine them into one buy-or-sell call.
How it works
A learned meta-family: for each ticker it assembles the per-family position matrix from the best passing champion of each family, then trains a gradient-boosted regressor to predict the ticker's next-day return from those family signals. Unlike fixed risk-balanced weighting/equal/consensus weights, it learns a non-linear combination of the families (a stacked-generalization-style meta-learner).
Data dependencies
- Strategy experiments
A data feed this strategy reads, refreshed on its normal schedule.
- Alpha family xs validation
A data feed this strategy reads, refreshed on its normal schedule.
- Daily prices
Adjusted-close OHLCV for every US-listed ticker; primary price feed.
Expected edge
Learns when each family is right and combines them non-linearly, capturing interactions a fixed-weight blend misses.
Related families
Explore learned stacked-ensemble family blend on alphactor.ai
See which tickers this family is currently firing on, with live signals and rankings.