Are you sure you want to delete this task? Once this task is deleted, it cannot be recovered.
You can not select more than 25 topics
Topics must start with a chinese character,a letter or number, can include dashes ('-') and can be up to 35 characters long.
tal
f7754c6a2c
|
1 year ago | |
|---|---|---|
| data | 1 year ago | |
| utils | 1 year ago | |
| LICENSE | 1 year ago | |
| README.md | 1 year ago | |
| side_info.py | 1 year ago | |
| train.py | 1 year ago | |
README.md
HIRE_0f34z57i8u
This is a tiny demo code for HIRE on movielens-100k dataset.
Requirements
- Python 3
- Scikit-learn
- Numpy
- Pandas
Training
In data file, training data has been splited with names u1.base, u2.base, u3.base, u4.base, u5.base.
Hierarchy matrix and flat feature matrix are available in .txt form in data folder.
All you need is to run train.py in terminal.
Evaluating
Test data has been defined with names u1.test, u2.test, u3.test, u4.test, u5.test in data folder.
The code will print RMSE value for test data with five fold cross-validation when you run train.py.
Dataset
The dataset is a copy of the MovieLens | GroupLens
dataset in the MovieLens 100k | GroupLens <http://files.grouplens.org/datasets/movielens/ml-100k.zip/>_
该算法提出了一个新的推荐系统框架用来解决异质性辅助信息进行推荐的问题，这些信息可以帮助理解用户和推荐的内容。该框架通过数学上的一致性来联合建模平面和层次的辅助信息。在3个真实场景数据集的实验结果证明了该框架的有效性。
Text Python other
