Skip to content

Latest commit

 

History

History
22 lines (12 loc) · 1007 Bytes

README.md

File metadata and controls

22 lines (12 loc) · 1007 Bytes
Raw

Demand Forecasting on Corporación Favorita Grocery Sales Data

Code

  1. Sampling_dataset.py: Sampling and preprocessing dataset

  2. Parameter_tunning_GBDT.py: Hyperparameter tunning on Gradient Boost Trees based model

  3. Parameter_tunning_DNN.py: Hyperparameter tunning on Deep Neural Network based model

  4. Train_GBDT.py: Training, validation and testing on Gradient Boost Trees based model

  5. Train_DNN.py: Training, validation and testing on Deep Neural Network based model

  6. Train_GBDT_perday.py: Training, validation and testing on Gradient Boost Trees based models for each day of the week

  7. Train_DNN_perday.py: Training, validation and testing on Deep Neural Network based models for each day of the week

  8. Visualise_Data.py: Daraw several visualisations used in the presentation

Data: obtained from kaggle

Required Python libraries: Numpy, Pandas, Sklearn, Tensorflow, Keras, Lightgbm, Matplotlib