K-Nearest Neighbors for Binary A-Share Return Classification

Summary

This brief strategy entry describes an A-share stock-selection example built around a k-nearest neighbors classifier. The accompanying discussion clarifies that the target is binary: a stock is labeled positive when its five-day return is above zero and negative otherwise. That framing illustrates how a trading model can turn a future-return condition into a supervised classification target.

The entry does not provide the model’s input features, neighborhood size, training procedure, validation design, trading rules, or performance results. A commenter questions whether the labeling expression creates only one class; the reply explains the intended positive-versus-nonpositive return split. Without further details, it is not possible to assess whether the implementation avoids look-ahead bias, handles class balance, or generalizes out of sample. The material is a minimal description of a machine-learning target rather than a reproducible strategy or evidence of profitability.

Key ideas

• The example applies k-nearest neighbors to A-share stock selection.
• Its intended target distinguishes positive five-day returns from nonpositive returns.
• The entry does not specify features, model settings, validation, or trading performance.
• The short discussion clarifies the binary label but does not establish model quality.

Tags

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