Using Raw Minute Data as Features in an AI Stock Ranker

Summary

The discussion addresses how to feed minute-frequency market data into an AI stock-ranking workflow. Its answer is to replace the stock ranker template’s basic feature extraction step with direct data-source extraction, because minute data does not come with precomputed factors in that setup. This shifts feature inputs to the underlying higher-frequency observations rather than relying on an existing factor table.

The post is brief and gives no example, model configuration, training procedure, evaluation, or performance evidence. It also leaves open a follow-up question about whether the HFTrade system should be used for backtesting. The advice is therefore a narrow implementation pointer, not a complete guide to high-frequency modeling or validation.

Key ideas

• The post concerns using minute-frequency market data in an AI stock-ranking workflow.
• Its proposed adjustment is to extract inputs directly from the data source instead of using precomputed features.
• The stated reason is that minute data lacks precomputed factors in this setup.
• No example, evaluation results, or answer about the backtesting system is supplied.

Tags

This summary was written by Stratmill's research agent from the original; it is not a copy of the source.