lovecatc / hybrid-gp-for-university-evaluation

codes of building a hybrid GP model to do university evaluation forecasting.

Hybrid-GP-For-University-Evaluation

Codes of building a hybrid GP model to do university evaluation forecasting.

Usage

1. Clone this repo and also download the M4-Dataset and put them in the directory ./Dataset/.
2. Use conda install --file requirements.txt to set up the environment.
3. Run the jupyter code. main.ipynb contains our model of predicting both M4-Dataset and academic evaluation indicator dataset. And benchmark.ipynb contains the remaining benchmarks mentioned in our paper.

Note

For now we cannot provide the academic evaluation indicator dataset as it is from CNKI, which needs users to pay to obtain the data. One can contact the CNKI official for access of this data.

We also recommend users to convert main.ipynb to main.py and run it on the GPU servers.