Round 10: Tossup 11

A form of this task that uses randomly-generated data is called “fuzzing.” In software engineering, code is written in a manner such that development (15[1])is “driven” by this task. Both this task and validation yield (-5[1])low accuracy (15[1]-5[1])for overfit models. In machine learning, twenty percent of data is typically reserved for this task rather than (15[1])training. The framework (*) JUnit is used to perform the “unit” form of this task on small pieces of code. This task often uses assert statements (10[1]-5[1])like assertTrue. Users manually do the “beta” form of this task for software pre-release. For 10 points, name this task of checking whether (10[1])code works. ■END■ (10[1])

Back to tossups

Buzzes

Summary