半监督学习基准数据集

半监督学习基准数据集

Semi-Supervised Learning Benchmark Dataset

该数据集出自:

Chapelle O, Scholkopf B, Zien A. Semi-supervised learning (chapelle, o. et al., eds.; 2006)[book reviews][J]. IEEE Transactions on Neural Networks, 2009, 20(3): 542-542.

网上链接为:

http://olivier.chapelle.cc/ssl-book/benchmarks.html

找了好久,网上个别链接打不开,存在此处方便以后查看。

另发现一个总结半监督数据集的好博客。

Self-Labeled Techniques for Semi-Supervised Learning: Taxonomy, Software and Empirical Study

https://sci2s.ugr.es/SelfLabeled

 

以下为第一个链接的内容:

The Benchmark Data Sets

For each data set, we provide 12 splits (exception: only 10 splits for data set 8) into labeled points and remaining unlabeled points. We ensure that each split contains at least one point of each class. Apart from this, there is no bias in the labeling process.

数据集存在12次划分,(但数据集8例外),每次有10/100个点为标记样本,其余为未标记样本,每次划分中每类至少一个点。

The table contains individual files in matlab 5.0 format (.mat files).

 

Data Set Points Dimensions Splits with l Labeled Points
g241c (set 5) 1500 241 l=10l=100
g241n (set 7) 1500 241 l=10l=100
Digit1 (set 1) 1500 241 l=10l=100
USPS (set 2) 1500 241 l=10l=100
COIL (set 6) 1500 241 l=10l=100
COIL2 (set 3) 1500 241 l=10l=100 (binary version of set 6)
BCI (set 4) 400 117 l=10l=100
Text (set 9) 1500 11960 l=10l=100
SecStr (set 8) 83,679 +
1,189,472
315 l=100l=1000l=10000
(no splitting of extra unlabeled data);
matlab script required

 

You can also download all data sets and splits (excluding the extra unlabeled data of set 8) at once as archive files, in matlab format: gzipped TAR fileZIP file; in ascii format: gzipped TAR fileZIP file (here, only the indices of the labeled examples are provided -- all other examples are unlabeled). Data sets 8 and 9 are supplied in special formats: in set 8, all attributes are categorical and have to be expanded into a sparse binary vector (21 bits per attribute; cf to the matlab script); in set 9, the data are very sparse, and only non-zero values are supplied as a list of "index:value" pairs.

X = matrix of input data; each row corresponds to one example

X输入数据的矩阵,每行对应于一个样本;
y = the labels (either {0,1} or {-1,+1} for binary problems)

y为标记,分{0,1}或{-1,+1}. 
idxLabs = each row contains the indices of the labeled points for a given split

idxLabs每行包含给定划分的标记样本点的索引
idxUnls = idem for the unlabeled points

idxUnls为未标记样本点
(all indices are 1-based as in matlab, not 0-based as in C)

索引从1开始。


Back to the SSL book main page

 

 

 

2020.3.3

posted @ 2020-03-03 22:36  上善若水啦  阅读(...)  评论(...编辑  收藏