# Multi-Class Cirrhosis Outcome Prediction — Data

来源：https://www.kaggle.com/competitions/cirrhosis-outcomes-challenge-a-multi-class-prediction-task/data

本地 Meta Kaggle 快照；不是本次实时页面。

# Dataset Description

The dataset for this competition was synthetically generated using a deep learning model trained on the **Cirrhosis Patient Survival Prediction** dataset. While the feature distributions closely resemble those of the original data, they are not identical, providing participants with a realistic yet slightly modified modeling challenge.

Competitors are encouraged to explore the relationship between the synthetic and original datasets. Incorporating the original data into your training pipeline may help improve generalization and predictive performance.

This competition focuses on predicting patient outcomes associated with liver cirrhosis — a progressive condition with varying prognoses depending on clinical factors and treatment pathways.

---

## **Files**

### **train.csv**

The training dataset contains the target variable `Status`, which represents the patient’s outcome:

* **C (Censored):** The patient was alive at `N_Days`.
* **CL (Censored due to Liver Transplant):** The patient was alive at `N_Days` following a liver transplant.
* **D (Deceased):** The patient had passed away by `N_Days`.

Use this dataset to train and validate your multi-class classification models.

---

### **test.csv**

The test dataset has the same features as the training data but does **not** include the target variable.

Your objective is to predict the probability for each possible outcome:

* `Status_C`
* `Status_CL`
* `Status_D`

---

### **sample_submission.csv**

A sample submission file demonstrating the correct structure and formatting required for leaderboard evaluation. Use this file as a reference to ensure your submission is properly formatted.