Blog โ€ข Getting Started

AI Model Classifications for Stock Research

A plain-English walkthrough for first-time users: what the model output means, how confidence should be interpreted, and how to use it as part of a better research workflow.

What the daily model classification is trying to do

The daily model classification is designed to help you organise your attention. It gives you a structured view of whether a ticker currently looks stronger, weaker, or more neutral based on the signals the model tracks.

It is not a prediction of guaranteed price movement. Think of it as a research prompt: it tells you where to look more closely, not what to do blindly.

How to read the confidence score

The confidence percentage reflects how strongly the current data pattern matches what the model has learned from similar conditions. A higher percentage means the model sees a clearer pattern. A lower percentage means the picture is more mixed.

Confidence is about pattern strength, not certainty. A 78% confidence score does not mean there is a 78% chance the stock will rise. It means the model has stronger conviction in its classification based on the information available.

What you should compare before reacting

  • Check whether the broader market is supportive or weak.
  • Review the ticker's recent price trend and volume behaviour.
  • Look for company announcements, earnings updates, or sector news.
  • Compare the classification with your own watchlist thesis and risk tolerance.

What not to assume

Do not treat a strong classification as a buy signal on its own. Do not assume a weak classification means a company is permanently poor quality. And do not ignore context such as macro news, liquidity, or sharp event-driven moves.

The model helps you prioritise research. Your judgment still matters โ€” very much so.

Want to see it in context?

Pair the daily classification with your watchlist and ticker detail review so the signal supports your thinking instead of replacing it.