Trading glossaryMetodologia

Bias da data mining

Data mining bias is finding patterns in historical data that appear profitable only by chance, because many rules or parameters were tested until one worked.

Explained

The more variations tested, the more likely a spurious winner. Defences include economic logic, out-of-sample testing, fewer parameters and adjusting for multiple tests.

Example

A trader tests 500 indicator combinations and picks the best backtest.

In live trading the strategy performs no better than random.

Common mistakes

01

Not counting how many tests were run.

02

Skipping out-of-sample checks.

03

Building strategies without a rationale.

Questions traders ask

How do I avoid data mining bias?

Limit tests, require logic and validate out of sample.

Is it the same as overfitting?

Closely related: overfitting tunes one model too closely to the data.

We (www.marketpedia.eu) and selected third parties (1) collect personal information as specified in the privacy policy and use cookies or similar technologies for technical purposes and, with your consent, for measurement and “marketing (personalized ads)” as specified in the cookie policy.

With respect to advertising, we and 1237 selected , may use precise geolocation data, and identification through device scanning in order to store and/or access information on a device and process personal data like your usage data for the following : personalised advertising and content, advertising and content measurement, audience research and services development.

You can freely give, deny, or withdraw your consent at any time by accessing the preferences panel. Your choices apply only to this domain. Denying consent may make related features unavailable.

Use the “Accept” button to consent. Use the “Reject” button to continue without accepting.