# mxnet examples with R
the purpose of this repo is simply to give you an easy access to mxnet and its api
## dependencies:
* mxnet
* ggplot2
* reshape2
* darch
## to use the code:
1. run preprocessing/get_mnist.R
2. run preprocessing/conversion.R
3. run models/
note that you have to edit directories in some files (to load data or save results)
## model overview
| Model | Time to execute* | Type |
| ------------- |:-------------------:|:---------------------------------------------------:|
| 01 | roughly 1.3 min | very simple (dense) network with 3 layers |
| 02 | roughly 43.5 min | 4 dense networks with dropout (batchsize benchmark |
| 03 | roughly 13.5 min | CNN with 3 conv + 3 dense layers |
| 04 | roughly 2 min | Denoising Autoencoder |
\* models were executed on mainstream CPU with 4 cores/4 threads @ 3.9 GHZ
I'm planing to add more models in the future (RNNs/lstm, image segmentation models)
## model 01 results:
![alt text](https://github.com/NiklasDL/mxnet-tutorials-in-R/blob/master/results/simpleNetErrors.png?raw=true)
## model 02 results:
![alt text](https://github.com/NiklasDL/mxnet-tutorials-in-R/blob/master/results/deepNetTrainError.png?raw=true)
![alt text](https://github.com/NiklasDL/mxnet-tutorials-in-R/blob/master/results/deepNetTestError.png?raw=true)
## model 03 results:
![alt text](https://github.com/NiklasDL/mxnet-tutorials-in-R/blob/master/results/cnnError.png?raw=true)
## model 04 results for arbitrary digits (top row shows original digits, intermediate row noised images used for training and bottom row prediction):
![alt text](https://github.com/NiklasDL/mxnet-tutorials-in-R/blob/master/results/denoising_autoencoder.png?raw=true)
好家伙VCC
- 粉丝: 1978
- 资源: 9140
最新资源
- androidx.multidex.MultiDexApplication.apk.1
- 丑子金装美化32(1).zip
- 基于Visual Basic .Net及Python技术的学校需求解决方案设计源码
- 基于Java语言的Web开发学习Demo设计源码
- 基于Java的医院排队叫号系统界面原型设计源码
- 基于Java语言的Himalaya仿喜马拉雅设计源码
- 基于PHP+HTML+CSS+JavaScript的智能电车管家设计源码
- 基于Emscripten编译的纯H5直播流播放器jessibuca设计源码
- 基于react-native的Android隐私合规配置与代码集成设计源码
- 基于JavaFX技术的全功能色彩管理器设计源码
资源上传下载、课程学习等过程中有任何疑问或建议,欢迎提出宝贵意见哦~我们会及时处理!
点击此处反馈