Need help interpreting research paper about Factorization Machines

进行中 已发布的 4 年前 货到付款
进行中 货到付款

There is a method of content recommendation called "Field Aware Factorization Machines" which is explained in this research paper: [login to view URL]~cjlin/papers/[login to view URL]

About this I need something very specific explained: The model.

Using an ffm implentation called xLearn ( [login to view URL] ) I have managed to train the model. However I don't understand how to interpret the trained model file. This file is about 8 mb large and is uploaded in the following link:

[login to view URL]

Fragments of this file are like so:

bias: -1.19493

i_0: 0.222636

i_1: 0

i_2: 0

i_3: 0

i_4: 0

i_5: 0

i_6: 0

i_7: 0

i_8: 0

i_9: 0

i_10: 0

i_11: 0

i_12: 0

.

.

.

v_9990_5: 0.163983 -0.00435618 0.205937 0.162927

v_9990_6: 0.0364918 0.181211 0.136226 0.0891592

v_9990_7: 0.0924703 0.307023 0.271298 0.156904

v_9990_8: 0.0628079 0.250727 0.0637604 0.294064

v_9990_9: 0.31339 0.3204 0.0064398 0.23125

v_9990_10: 0.123403 0.323897 0.200116 0.22379

v_9990_11: 0.0808216 0.32948 0.0250665 0.257791

v_9990_12: 0.322103 0.0737792 0.0105526 0.293231

v_9990_13: 0.315756 0.298412 0.310376 0.0305769

v_9990_14: 0.0390615 0.0692963 0.019608 0.145432

v_9990_15: 0.119959 0.0367788 0.254127 0.0489978

v_9990_16: 0.100716 0.216424 0.00206306 0.091204

v_9990_17: -0.00010065 0.19462 0.120955 0.0980957

I need someone to explain to me what this values are and how are they used. Please interpret the research paper and show me how to understand this.

Thanks!

Michel

机器学习(ML)

项目ID: #19291768

关于项目

远程项目 活跃的4 年前