There is a newer version of the record available.

Published June 29, 2023 | Version 1

MNIST-Federated-Learning

  • 1. LISTIC Laboratory Université Of Savoie Mont Blance

Description

Please find below the descriptions of the three configurations for partitioning the MNIST Train dataset into 10 clients and the MNIST Train data: 
 

  1. Balanced Distribution: In the first configuration, the MNIST dataset is partitioned among 10 clients in a balanced manner. This means that the data samples from each class are evenly distributed among the clients. Each client receives a roughly equal number of images from each digit class, ensuring that the distribution of samples across clients is proportional and representative of the overall dataset.    [ Config 1]
  2. Heterogeneous Distribution (One Class per Client): In the second configuration, the MNIST dataset is partitioned in a heterogeneous manner, where each client is assigned a single digit class exclusively. This means that one client will only receive images of the digit '0', another client will receive images of the digit '1', and so on. In this setup, each client becomes an expert in classifying a specific digit, allowing for specialized training and evaluation. [ Config 2]
  3. Mixed Distribution: In the third configuration, the MNIST dataset is partitioned using a mixed distribution approach. This means that the data samples from all digit classes are distributed among the 10 clients, but the distribution is not necessarily balanced. The number of samples assigned to each client may vary for different digit classes, resulting in an uneven distribution across the clients. This configuration aims to capture both the overall diversity of the dataset and the varying difficulty levels of classifying different digits. [ Config 3 ]

 

The structure of "Mnist-dataset" folder is :
Mnist-dataset/
├── config1/
│   ├── client-1/
│   │   └── client_1_config1.csv
│   ├── client-2/
│   │   └── client_2_config1.csv
│   ├── client-3/
│   │   └── client_3_config1.csv
│   └── ...
├── config2/
│   ├── client-1/
│   │   └── client_1_config2.csv
│   ├── client-2/
│   │   └── client_2_config2.csv
│   ├── client-3/
│   │   └── client_3_config2.csv
│   └── ...
├── config3/
│   ├── client-1/
│   │   └── client_1_config3.csv
│   ├── client-2/
│   │   └── client_2_config3.csv
│   ├── client-3/
│   │   └── client_3_config3.csv
│   └── ...
└── mnist_test.csv
 

Files

client10_config3.csv

Files (441.1 MB)

Name Size
md5:dc63452fd39d54496fb776271a8f527a
55.9 MB Preview Download
md5:9a2d6429b37390cb05790edc7f175030
33.5 MB Preview Download
md5:bf0c5e7633952c76d1b3d805edd32755
10.9 MB Preview Download
md5:b29501aa388057c83789867503bac0a6
20.8 MB Preview Download
md5:bbfd8928c13eb63c120e6836043e667f
10.8 MB Preview Download
md5:8987c070b83e13a088d7df3cd424a68f
42.7 MB Preview Download
md5:c2575a57a59706cc14a6a0a5d9fb5980
9.9 MB Preview Download
md5:4041d0dc4690b7227725e3e75325683a
22.7 MB Preview Download
md5:ccb37324bec350bef1020d0d17f26f3c
11.0 MB Preview Download
md5:393e7cc2a04cc3047c85111e1d740365
10.6 MB Preview Download
md5:ecd4f8aba608090a0598683c3190eaaf
10.8 MB Preview Download
md5:893cc3283fc38b73fa6a6d5444c1688c
11.0 MB Preview Download
md5:50708f72bdc6194096af7ef3a0c66205
11.1 MB Preview Download
md5:2c9b32fc3c3a5e80b9f711b37f49bbba
22.2 MB Preview Download
md5:f5506458e45f74a52b43e09a7266345f
11.0 MB Preview Download
md5:e2c369e8b3631b7b568a0edacfd93af6
10.9 MB Preview Download
md5:6dcf57f4f97c1022ab5f5c847c250af2
22.7 MB Preview Download
md5:faa175eecd7c771bfc82cfbf3c45c46f
11.0 MB Preview Download
md5:026cd2fb517dea19438a1ecdfc01ff76
11.6 MB Preview Download
md5:015463cffa036aa67a4c50a0a86c2249
11.3 MB Preview Download
md5:822a17fd688b83b6f25df05b91f984bd
60.8 MB Preview Download
md5:79e4062bbe5540ed681119c464256ebb
18.3 MB Preview Download

Additional details

References

  • DENG, Li. The mnist database of handwritten digit images for machine learning research [best of the web]. IEEE signal processing magazine, 2012, vol. 29, no 6, p. 141-142.