Statistical Testing

To produce valid, reliable, and reproducible results, it is important to ensure that a discovered pattern or relationship in the data is not an outcome of random chance, but rather represents a significant effect that did not occur due to natural variability in the data samples. This often involves using statistical procedures for avoiding false discoveries. A statistical test is a generic procedure for measuring the evidence for accepting or rejecting a hypothesis that the outcome (result) of an experiment or a data analysis procedure provides.

Find a recent scholarly article involving statistical testing techniques in data mining from a credible source on the Internet or the UC Library (no blogs, no book chapters).

In your initial post, please describe the technique and the significance of the procedure in the realm of data mining and validity testing.

Here is an example of an article: Instance-Based Classification through Hypothesis Testing

To earn credit, the article must be published within the past three years. You must provide a direct link to the article for verification purposes.

Respond substantively to at least two other students’ posts. In your peer responses, discuss examples of how this technique would be used to solve data mining issues that many data scientists face today.

*Please remember to include at least one credible scholarly reference with your initial post!