# DataMining_2025Spring_Homework2 — Data

来源：https://www.kaggle.com/competitions/data-mining-hw-2/data

采集时间：2026-10-18T13:59:20Z


The dataset is provided in the [AnnData](https://anndata.readthedocs.io/en/stable/index.html) (`.h5ad`) format, a widely used data structure for single-cell data. In this format:
- Rows represent cells, and columns represent features (i.e., genome bins).
- The batch annotations for each cell are stored in `adata.obs['batch']`. Participants can exlpore effective batch effect removal method to handle batch effects in data.
- The dataset contains NaN values. Participants should take appropriate measures to handle these missing values effectively. 
## Files

*   **final_dataset.h5ad** - the dataset with batch annotations
*   **sample_submission.csv** - a sample submission file in the correct format
For each ID in the dataset, you must assign a cluster for the TARGET variable. The file should contain a header and have the following format:
```python
    ID,TARGET
    0, 0
    1, 1
    2, 1
```
    etc, where ID is the obs_names of single-cell data.