Match-Target — Reverse-Engineer the Credit Decision Rule

You're a credit analyst handed 12 customer profiles and a column showing the expected decision per customer. The bank's underwriting model is documented somewhere, but you don't have it — you need to derive the rule from the target column itself, then encode it in a formula that exactly reproduces the targets.

Columns:

• A — Customer
• B — Credit Score (300–850 scale)
• C — Annual Income
• D — Months Employed at current job
• E — Your Decision (you fill this with a formula)
• F — Target Decision (the source of truth — what you're trying to match)

The grader compares column E against the spec — column F is your sanity-check.

By inspecting the data carefully, you should be able to derive a 5-rule precedence ladder using credit score and income (months employed turns out to be a red herring — it's there to test whether you over-fit). The hint sheet lists the rule structure without naming the thresholds:

1. Top credit score → Approve (no other check)
2. Solid credit score AND solid income → Approve
3. Bad credit score → Decline
4. Solid credit score BUT inadequate income → Decline (low-income screen)
5. Everything else → Manual Review

Your task:

1. Inspect column F to identify the threshold values that separate "top", "solid", and "bad" credit scores, and the income cutoffs.
2. Write an IF formula in column E that produces the same decisions as column F for every row.

Graded cells: E3 (Globex), E5 (Hooli), E7 (Vandelay).