可信智能计算服务 TICS-准备数据

时间：2024-04-02 16:37:49

准备数据

首先，企业A和大数据厂商B需要商议确定要提供的数据范围及对应的元数据信息，例如双方初始决定使用最近三个月的已有用户转化数据作为联邦训练的训练集和评估集。

表1 企业A的数据
字段名称	字段类型	描述
id	string	hash过后的手机号字符串
col0-col4	float	企业A数据特征
label	int	企业A对用户的标签属性

industry1.csv

id,col0,col1,col2,col3,col4,label
19581e27de7ced00ff1ce50b2047e7a567c76b1cbaebabe5ef03f7c3017bb5b7,-0.823913755,0.787712038,0.429635596,-1.315646486,-1.652321611,1
2c624232cdd221771294dfbb310aca000a0df6ac8b66b696d90ef06fdefb64a3,3.041881096,-0.651684341,3.661649955,0.035548734,3.477873904,0
3fdba35f04dc8c462986c992bcf875546257113072a909c162f7e470e581e278,-1.847252571,0.496981447,1.654416521,-1.945006902,0.394151993,1
4523540f1504cd17100c4835e85b7eefd49911580f8efff0599a8f283be6b9e3,-0.593556893,-0.351750558,0.964512256,-0.017390132,0.092562565,1
4a44dc15364204a80fe80e9039455cc1608281820fe2b24f1e5233ade6af1dd5,0.241505219,-0.219114719,1.51438745,-0.665234511,0.178575706,0
4b227777d4dd1fc61c6f884f48641d02b4d121d3fd328cb08b5531fcacdabf8a,0.372607556,-0.29194018,0.080862655,0.391501604,-0.012276428,1
4e07408562bedb8b60ce05c1decfe3ad16b72230967de01f640b7e4729b49fce,1.544282251,-0.203027285,3.076050022,-0.530666302,2.156693386,0
4ec9599fc203d176a301536c2e091a19bc852759b255bd6818810a42c5fed14a,1.006651366,-0.972403786,1.314115256,0.363296291,5.171128738,0
4fc82b26aecb47d2868c4efbe3581732a3e7cbcc6c2efb32062c08170a05eeb8,-2.859681221,-1.465959913,-0.930994729,-0.773533542,-3.673734138,0
5feceb66ffc86f38d952786c6d696c79c2dbc239dd4e91b46729d73a27fb57e9,-1.409250598,-0.589367921,-4.467396693,1.370376188,-1.2368325,1

大数据厂商B的数据如下，一共有10条记录。

表2 大数据厂商B的数据
字段名称	字段类型	描述
id	string	hash过后的手机号字符串
f0-f4	float	大数据厂商数据特征

bigdata1.csv

id,f0,f1,f2,f3,f4
2c624232cdd221771294dfbb310aca000a0df6ac8b66b696d90ef06fdefb64a3,0.390064223,0.664175034,3.20228741,0.380574513,0.017733811
3fdba35f04dc8c462986c992bcf875546257113072a909c162f7e470e581e278,-0.483250226,0.616586578,3.001851708,2.407914633,0.856369412
4a44dc15364204a80fe80e9039455cc1608281820fe2b24f1e5233ade6af1dd5,-0.070919538,-2.219653517,1.461645551,1.66185096,0.778770954
4ec9599fc203d176a301536c2e091a19bc852759b255bd6818810a42c5fed14a,0.024227451,-1.087235302,3.67470964,-2.420729037,-3.132456573
4fc82b26aecb47d2868c4efbe3581732a3e7cbcc6c2efb32062c08170a05eeb8,-0.771151327,-1.184821181,-0.674077615,-0.379858223,0.158957184
6b51d431df5d7f141cbececcf79edf3dd861c3b4069f0b11661a3eefacbba918,-0.738091802,-1.474822882,2.93475295,-3.763763721,-1.817301398
6b86b273ff34fce19d6b804eff5a3f5747ada4eaa22f1d49c01e52ddb7875b4b,-1.216062821,-1.093614452,-1.632396806,0.887601314,-4.40930101
8527a891e224136950ff32ca212b45bc93f69fbb801c3b1ebedac52775f99e61,-0.789268594,1.071733834,3.763254446,-3.760298263,0.49776472
e7f6c011776e8db7cd330b54174fd76f7d0216b612387a5ffcfb81e6f0919683,-2.759963795,0.405262468,1.264947591,1.027350049,1.293868423

其中为了保证数据安全，企业A和大数据厂商B通过讨论决定使用hash过后的手机号作为已有数据的唯一标识id字段，并将唯一标识作为数据对齐的依据。

父主题： 使用 TICS 多方安全计算进行联合样本分布统计

上一篇：可信智能计算服务 TICS-场景描述

下一篇：可信智能计算服务 TICS-场景描述