# 2.28.22:🦠Bacteria — Data

来源：https://www.kaggle.com/competitions/22822bacteria/data

本地 Meta Kaggle 快照；不是本次实时页面。

# Dataset Description

See (Tabular Playground Series - Feb 2022](https://www.kaggle.com/c/tabular-playground-series-feb-2022/data) for the full description of this dataset. We use the same number of (augmented) features, only 5 bacteria classes and 100K observations, of which the first 10K are designated for Kaggle testing purpose. You have 90K of the remaining observations to train your models on. The output variable is labeled with IDs 0, 1, 2, 3, 4.

The original features were representing the lossy scaled and shifted histogram (i.e. empirical probabilities) of the base counts after subtracting some random base. The features in this competition were further augmented in some non uniformly random fashion.

Here are the bacteria are categorizing (not necessarily in this order or with these integer IDs):

1. Bacteroides_fragilis
1. Campylobacter_jejuni
1. Salmonella_enterica
1. Streptococcus_pneumoniae
1. Streptococcus_pyogenes

Good luck!