https://github.com/sdobber/FluxArchitectures.jl
Complex neural network examples for Flux.jl
https://github.com/sdobber/FluxArchitectures.jl
Last synced: 2 months ago
JSON representation
Complex neural network examples for Flux.jl
- Host: GitHub
- URL: https://github.com/sdobber/FluxArchitectures.jl
- Owner: sdobber
- License: mit
- Created: 2020-04-05T16:50:26.000Z (over 5 years ago)
- Default Branch: master
- Last Pushed: 2023-11-22T17:13:39.000Z (over 1 year ago)
- Last Synced: 2024-08-03T20:04:24.723Z (11 months ago)
- Language: Julia
- Size: 44.3 MB
- Stars: 122
- Watchers: 12
- Forks: 15
- Open Issues: 10
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
- awesome-sciml - sdobber/FluxArchitectures.jl: Complex neural network examples for Flux.jl
README
# FluxArchitectures
[](https://sdobber.github.io/FluxArchitectures.jl/dev)
[](https://github.com/sdobber/FluxArchitectures.jl/actions)
[](https://codecov.io/gh/sdobber/FluxArchitectures.jl)
Complex neural network examples for Flux.jl.
This package contains a loose collection of (slightly) more advanced neural network architectures, mostly centered around time series forecasting.
## Installation
To install FluxArchitectures, type `]` to activate the package manager, and type
```julia
add FluxArchitectures
```
for installation. After `using FluxArchitectures`, the following functions are exported:
* `prepare_data`
* `get_data`
* `DARNN`
* `DSANet`
* `LSTnet`
* `TPALSTM`
See their docstrings, the [documentation]((https://sdobber.github.io/FluxArchitectures.jl/stable)), and the `examples` folder for details.
## Models
* **LSTnet**: This "Long- and Short-term Time-series network" follows the paper by [Lai et. al.](https://arxiv.org/abs/1703.07015).
* **DARNN**: The "Dual-Stage Attention-Based Recurrent Neural Network for Time Series Prediction" is based on the paper by [Qin et. al.](https://arxiv.org/abs/1704.02971).
* **TPA-LSTM**: The Temporal Pattern Attention LSTM network is based on the paper "Temporal Pattern Attention for Multivariate Time Series Forecasting" by [Shih et. al.](https://arxiv.org/pdf/1809.04206v2.pdf).
* **DSANet**: The "Dual Self-Attention Network for Multivariate Time Series Forecasting" is based on the paper by [Siteng Huang et. al.](https://kyonhuang.top/files/Huang-DSANet.pdf)
## Quickstart
Activate the package and load some sample-data:
```julia
using FluxArchitectures
poollength = 10; horizon = 15; datalength = 1000;
input, target = get_data(:exchange_rate, poollength, datalength, horizon)
```
Define a model and a loss function:
```julia
model = LSTnet(size(input, 1), 2, 3, poollength, 120)
loss(x, y) = Flux.mse(model(x), y')
```
Train the model:
```julia
Flux.train!(loss, Flux.params(model),Iterators.repeated((input, target), 20), Adam(0.01))
```