DeepTUL

PyTorch implementation of AAMAS'20 paper-Trajectory-User Linking with Attentive Recurrent Network

Datasets

The sample data to evaluate our model can be found in the data folder, which contains 200+ users and ready for directly used.

Requirements

• Python 2.7
• Pytorch 0.4.1

cPickle is used in the project to store the preprocessed data and parameters.

Project Structure

• /codes
  • main.py
  • model.py # define models
  • masked_cross_entropy.py #calculate entropy
  • train.py # define tools for train the model
  • /data
• /data # preprocessed foursquare sample data (pickle file)
• /docs # paper and presentation file
• /resutls # the default save path when training the model

Usage

Train a new model:

python main.py 

Other parameters (refer to main.py):

• for training:
  • learning_rate, lr_step, lr_decay, L2, clip, epoch_max, dropout_p
• model definition:
  • loc_emb_size, uid_emb_size, tim_emb_size, hidden_size, rnn_type, attn_type
  • strategies_type: AVE-sdot,AVE-dot,MAX-sdot,MAX-dot

Other

More specific data and data processing methods will be given later