The task
Given one column of messy real values and a handful of worked examples, produce the transformation that splits or normalises the whole column.
Proposed by PIG-DemoRank 10 of 12Not in the first set
Data Column Split
Who it is for
Chief Data Officer
A large part of my analytics backlog is a person reshaping a spreadsheet by hand and calling it a project.
The reward
Pays exact match on held-out rows the model never saw while it was writing the rule.
The counterweight
The rule is applied to those rows, not fitted to them, and a rule that special-cases individual values is penalised on length. Memorising the examples scores zero on the rows that pay.
The counterweight is the part worth arguing about. A reward with nothing pulling against it is a target, and a target is what a model learns to hit instead of doing the job.
What exists today
Nothing. There is no environment for this vertical, no recorded run, and no number. It is a proposal, written to the same shape as the one that is built so you can judge it on the same terms.
The environment that does exist is Word Five. It is a word game, and it is deliberately not one of these twelve: it is there to show the machinery, not to imply we have built yours.