MoST

This repository contains the official implementation for the paper []

Requirements

The recommended requirements for MoST are specified as follows:

• Python 3.8.12
• torch==1.11
• scipy==1.6.1
• numpy==1.24.2
• pandas==2.0.0
• scikit_learn==0.24.2
• statsmodels==0.12.2
• Bottleneck==1.3.2

The dependencies can be installed by:

pip install -r requirements.txt

Data

The datasets can be obtained and put into datasets/ folder in the following way:

• [Google trend datasets] is put into datasets/country or datasets/region so that each data file can be located by datasets/country/<ID>/<query>.csv.
• KnowAir datasets should be put into datasets/ so that each data file can be located by datasets/KnowAir.npy.

Usage

To train and evaluate MoST on a dataset, run the following command:

python train.py <dataset_name> <run_name> --loader <loader> --batch-size <batch_size> --max-train-length <max_train_length> --repr-dims <repr_dims> --gpu <gpu> --epochs <epochs> --eval

python train.py e_commerce e_commerce --loader forecast_tensor --batch-size 8 --max-train-length 200 --repr-dims 320 --gpu 0 --epochs 100 --eval --seed 1

The detailed descriptions about the arguments are as following:

Parameter name	Description of parameter
dataset_name	The dataset name
run_name	The folder name used to save model, output and evaluation metrics. This can be set to any word
loader	The data loader used to load the experimental data. This can be set to forecast_tensor or classification_tensor or encode_tensor
batch_size	The batch size (defaults to 8)
max_train_length	The size of lookback window (defaults to 200)
repr_dims	The representation dimensions (defaults to 320)
gpu	The gpu no. used for training and inference (defaults to 0)
eval	Whether to perform evaluation after training

After training and evaluation, the trained encoder, output and evaluation metrics can be found in training/DatasetName__RunName_Date_Time/ and result.

Scripts: The scripts for reproduction are provided in scripts/ folder.