# Anomaly detection — Data

来源：https://www.kaggle.com/competitions/eecs498/data

本地 Meta Kaggle 快照；不是本次实时页面。

# Dataset Description


## File descriptions

*   **conn250K.csv** - this file is in the same format as "conn250K.csv" as you have seen in the handout of project 2 -- in fact, it was recorded separately for the same host described in the handout. There are totally 256670 records, each of which is with 4 fields that are described in the "Data fields" section. You will run the outlier detection technique you implemented on this dataset and upload the detection result on Kaggle.


## Data fields

*   **record ID** - The unique identifier for each connection record.
*   **duration**_  This feature denotes the number of seconds (rounded) of the connection. For example, a connection for 0.17s or 0.3s would be indicated with a “0” in this field.
*   **src\_bytes** This field represents the number of data bytes transferred from the source to the destination (i.e., the amount of out-going bytes from the host).
*   **dst\_bytes** This feature represents the number of data bytes transferred from the destination to the source (i.e., the amount of bytes received by the host).

## What to submit
Your detection result should be in the same format as described in the handout of project 2. Specifically, there should be only 2 columns separated by the comma:

*   **record ID** - The unique identifier for each connection record.
*   **is\_anomaly?**_  This binary field indicates your detection result: **0** denotes the transmission is normal, **1** indicates anomalous.

## How is your result evaluated
Refer to the "Evaluation" section under "overview".