https://github.com/ivanbongiorni/rnn_timeseries-forecast

An LSTM Neural Network for Time Series Forecasting, trained on Wikipedia's Web Traffic dataset from Kaggle.
https://github.com/ivanbongiorni/rnn_timeseries-forecast

dataset forecasting kaggle recurrent-neural-network rnn series-forecasting tensorflow tensorflow2 time-series timeseries-analysis timeseries-forecasting web-traffic wikipedia

Last synced: 14 days ago
JSON representation

An LSTM Neural Network for Time Series Forecasting, trained on Wikipedia's Web Traffic dataset from Kaggle.

• Host: GitHub
• URL: https://github.com/ivanbongiorni/rnn_timeseries-forecast
• Owner: IvanBongiorni
• License: mit
• Created: 2020-03-27T17:06:03.000Z (over 4 years ago)
• Default Branch: master
• Last Pushed: 2020-11-21T14:18:27.000Z (almost 4 years ago)
• Last Synced: 2024-10-10T17:23:08.442Z (27 days ago)
• Topics: dataset, forecasting, kaggle, recurrent-neural-network, rnn, series-forecasting, tensorflow, tensorflow2, time-series, timeseries-analysis, timeseries-forecasting, web-traffic, wikipedia
• Language: Jupyter Notebook
• Homepage:
• Size: 5.58 MB
• Stars: 1
• Watchers: 1
• Forks: 2
• Open Issues: 0
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

        
# Recurrent Neural Network for Time Series Forecasting



---

This is a time series forecasting project based on the Wikipedia [Web Traffic Time Series Forecasting](https://www.kaggle.com/c/web-traffic-time-series-forecasting) dataset from Kaggle. 

Two **RNN architectures** are implemented:

- A "Vanilla" RNN regressor.

- A Seq2seq regressor.

Both are implemented in **TensorFlow 2**, with *custom training functions* optimized with **Autograph**.

## Structure of the repository

Main files:

- `config.yaml`: config file for hyperparameters.

- `dataprep.py`: data preprocessing pipeline.

- `train.py`: training pipeline.

- `tools.py`: contains useful processing functions to be iterated in main pipelines.

- `model.py`: builds model.

I also added a `visualize_performance.ipynb` Jupyter Notebook to visually inspect models' performance on Test data.

Folders:

- `/data_raw/`: requires unzipped `train_2.csv` file from [Kaggle](https://www.kaggle.com/c/web-traffic-time-series-forecasting/). Available is an `imputed.csv` dataset, containing imputed time series, coming from my other repository on a [GAN for imputation of missing data in time series](https://github.com/IvanBongiorni/GAN-RNN_Timeseries-imputation).

- `/data_processed/`: divided in `/Train/` and `/Test/` directories.

- `/saved_models/`: contains all saved TensorFlow models, both regressors.

- `/utils/`: for pics and other secondary files.

## How to run code

After you clone the repository locally, download the raw dataset from [Kaggle](https://www.kaggle.com/c/web-traffic-time-series-forecasting/), and place unzipped `train_2.csv` file in `/data_raw/` folder.

Then, time series forecast is executed in two steps. First, run data preprocessing pipeline:

`python -m dataprep`

This will generate Training+Validation and Test files, stored in `/data_processed/` subdirectories. Second, launch training pipeline with:

`python -m train`

This will either create, train and save a new model, or load and train an already existing one, stored in `/saved_models/` folder.

Finally, Test set performance will be evaluated from `test.ipynb` notebook.

## Modules

```

numpy==1.18.3

pandas==1.0.3

scikit-learn==0.22.2.post1

scipy==1.4.1

tensorflow==2.1.0

tqdm==4.45.0

```

## Hardware

I used a pretty powerful laptop, with 64GB or RAM and NVidia RTX 2070 GPU. I highly recommend GPU training to avoid excessive computational times.