Awesome

Reproduce performance of NBeats in M4 Dataset

Link to [paper].

Steps to reproduce

• Download M4 Dataset at Mcompetitions/M4-methods
• Placed it at somewhere and modifiled global variable DATA_PATH in data.py properly, the directory structure is:

  Dataset
  ├── M4-info.csv
  ├── Test
  │   ├── Daily-test.csv
  │   ├── Hourly-test.csv
  │   ├── Monthly-test.csv
  │   ├── Quarterly-test.csv
  │   ├── Weekly-test.csv
  │   └── Yearly-test.csv
  └── Train
      ├── Daily-train.csv
      ├── Hourly-train.csv
      ├── Monthly-train.csv
      ├── Quarterly-train.csv
      ├── Weekly-train.csv
      └── Yearly-train.csv
  

• Install missing packages like tensorflow_io
• Run the .ipynb file.

Results

There is only results for Yearly dataset and interpretable model at the moment. Check author's results here

Credit:

• philipperemy/n-beats
• ElementAI/N-BEATS
• Mcompetitions/M4-methods