kddcup99

• Description:

This is the data set used for The Third International Knowledge Discovery and Data Mining Tools Competition, which was held in conjunction with KDD-99 The Fifth International Conference on Knowledge Discovery and Data Mining. The competition task was to build a network intrusion detector, a predictive model capable of distinguishing between 'bad' connections, called intrusions or attacks, and 'good' normal connections. This database contains a standard set of data to be audited, which includes a wide variety of intrusions simulated in a military network environment.

• Additional Documentation: Explore on Papers With Code

• Homepage: https://kdd.ics.uci.edu/databases/kddcup99/kddcup99.html

• Source code: tfds.datasets.kddcup99.Builder

• Versions:

  • 1.0.0: Initial release.
  • 1.0.1 (default): Fixes parsing of boolean fields land, logged_in, root_shell, is_hot_login and is_guest_login.

• Download size: 18.62 MiB

• Dataset size: 5.25 GiB

• Auto-cached (documentation): No

• Splits:

Split	Examples
'test'	311,029
'train'	4,898,431

• Feature structure:

FeaturesDict({
    'count': int32,
    'diff_srv_rate': float32,
    'dst_bytes': int32,
    'dst_host_count': int32,
    'dst_host_diff_srv_rate': float32,
    'dst_host_rerror_rate': float32,
    'dst_host_same_src_port_rate': float32,
    'dst_host_same_srv_rate': float32,
    'dst_host_serror_rate': float32,
    'dst_host_srv_count': int32,
    'dst_host_srv_diff_host_rate': float32,
    'dst_host_srv_rerror_rate': float32,
    'dst_host_srv_serror_rate': float32,
    'duration': int32,
    'flag': ClassLabel(shape=(), dtype=int64, num_classes=11),
    'hot': int32,
    'is_guest_login': bool,
    'is_hot_login': bool,
    'label': ClassLabel(shape=(), dtype=int64, num_classes=40),
    'land': bool,
    'logged_in': bool,
    'num_access_files': int32,
    'num_compromised': int32,
    'num_failed_logins': int32,
    'num_file_creations': int32,
    'num_outbound_cmds': int32,
    'num_root': int32,
    'num_shells': int32,
    'protocol_type': ClassLabel(shape=(), dtype=int64, num_classes=3),
    'rerror_rate': float32,
    'root_shell': bool,
    'same_srv_rate': float32,
    'serror_rate': float32,
    'service': ClassLabel(shape=(), dtype=int64, num_classes=71),
    'src_bytes': int32,
    'srv_count': int32,
    'srv_diff_host_rate': float32,
    'srv_rerror_rate': float32,
    'srv_serror_rate': float32,
    'su_attempted': int32,
    'urgent': int32,
    'wrong_fragment': int32,
})

• Feature documentation:

Feature	Class	Shape	Dtype	Description
	FeaturesDict
count	Tensor		int32
diff_srv_rate	Tensor		float32
dst_bytes	Tensor		int32
dst_host_count	Tensor		int32
dst_host_diff_srv_rate	Tensor		float32
dst_host_rerror_rate	Tensor		float32
dst_host_same_src_port_rate	Tensor		float32
dst_host_same_srv_rate	Tensor		float32
dst_host_serror_rate	Tensor		float32
dst_host_srv_count	Tensor		int32
dst_host_srv_diff_host_rate	Tensor		float32
dst_host_srv_rerror_rate	Tensor		float32
dst_host_srv_serror_rate	Tensor		float32
duration	Tensor		int32
flag	ClassLabel		int64
hot	Tensor		int32
is_guest_login	Tensor		bool
is_hot_login	Tensor		bool
label	ClassLabel		int64
land	Tensor		bool
logged_in	Tensor		bool
num_access_files	Tensor		int32
num_compromised	Tensor		int32
num_failed_logins	Tensor		int32
num_file_creations	Tensor		int32
num_outbound_cmds	Tensor		int32
num_root	Tensor		int32
num_shells	Tensor		int32
protocol_type	ClassLabel		int64
rerror_rate	Tensor		float32
root_shell	Tensor		bool
same_srv_rate	Tensor		float32
serror_rate	Tensor		float32
service	ClassLabel		int64
src_bytes	Tensor		int32
srv_count	Tensor		int32
srv_diff_host_rate	Tensor		float32
srv_rerror_rate	Tensor		float32
srv_serror_rate	Tensor		float32
su_attempted	Tensor		int32
urgent	Tensor		int32
wrong_fragment	Tensor		int32

• Supervised keys (See as_supervised doc): None

• Figure (tfds.show_examples): Not supported.

• Examples (tfds.as_dataframe):

• Citation:

@misc{Dua:2019 ,
  author = "Dua, Dheeru and Graff, Casey",
  year = 2017,
  title = "{UCI} Machine Learning Repository",
  url = "http://archive.ics.uci.edu/ml",
  institution = "University of California, Irvine, School of Information and
Computer Sciences"
}