This competition provides an **anonymised, synthetic clinical dataset**. Each row represents a unique patient encounter; each of the 64 features is a binary flag indicating the presence (`1`) or absence (`0`) of a specific sign, symptom, diagnostic code, or laboratory threshold breach.
Your goal is to map these indicators to the correct disease category in `label`.

#### Files

| File                       | Shape    | Contents                                |
| -------------------------- | -------- | --------------------------------------- |
| **train.csv**              | 564 × 66 | 64 features + `label` + ID              |
| **test.csv**               | 143 × 65 | 64 features + ID (no `label`)           |
| **sample\_submission.csv** | 143 × 2  | `ID` and dummy `label`; use as template |

The public leaderboard score is based on **50%** of the hidden labels; the final private score (revealed at competition end) uses the remaining 50% so your rank in the leaderboard might be a little different after competition ends.