{"id":16021429,"url":"https://github.com/tentone/semeionnet","last_synced_at":"2025-04-10T01:34:38.752Z","repository":{"id":97242850,"uuid":"106745340","full_name":"tentone/semeionNet","owner":"tentone","description":"A set of machine learning experiments with the semeion and MNIST handwritten digit dataset using tensorflow","archived":false,"fork":false,"pushed_at":"2020-03-15T17:01:19.000Z","size":11408,"stargazers_count":3,"open_issues_count":0,"forks_count":4,"subscribers_count":2,"default_branch":"master","last_synced_at":"2025-03-24T03:12:13.674Z","etag":null,"topics":["mnist","semeion","tensorflow"],"latest_commit_sha":null,"homepage":null,"language":"Python","has_issues":false,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":"mit","status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/tentone.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":null,"funding":null,"license":"LICENSE","code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":null,"support":null,"governance":null,"roadmap":null,"authors":null,"dei":null,"publiccode":null,"codemeta":null}},"created_at":"2017-10-12T21:05:35.000Z","updated_at":"2022-05-09T13:35:04.000Z","dependencies_parsed_at":"2023-03-12T21:00:11.615Z","dependency_job_id":null,"html_url":"https://github.com/tentone/semeionNet","commit_stats":{"total_commits":7,"total_committers":1,"mean_commits":7.0,"dds":0.0,"last_synced_commit":"3d287c40ee9a7cc37c39f2575abe2b0018781668"},"previous_names":[],"tags_count":0,"template":false,"template_full_name":null,"repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/tentone%2FsemeionNet","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/tentone%2FsemeionNet/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/tentone%2FsemeionNet/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/tentone%2FsemeionNet/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/tentone","download_url":"https://codeload.github.com/tentone/semeionNet/tar.gz/refs/heads/master","host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":248140939,"owners_count":21054371,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2022-07-04T15:15:14.044Z","host_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub","repositories_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories","repository_names_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repository_names","owners_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners"}},"keywords":["mnist","semeion","tensorflow"],"created_at":"2024-10-08T18:04:14.445Z","updated_at":"2025-04-10T01:34:38.724Z","avatar_url":"https://github.com/tentone.png","language":"Python","funding_links":[],"categories":[],"sub_categories":[],"readme":"# SemeionNet\n\n - A set of machine learning experiments with the semeion and MNIST handwritten digit dataset using  tensorflow\n\n - The objective of this experiment was to test multiple classification methods using the semeion handwriting dataset and measure performance of different classifiers implementation in tensorflow. \n\n   \n\n### Dataset\n - The semeion dataset is composed of 1593 handwritten digits from 80 persons that were scanned and stretched to a 16x16 size image.\n\n  - http://archive.ics.uci.edu/ml/datasets/semeion+handwritten+digit\n\n - MNIST dataset is a subset of the NIST dataset that has over 60000 handwritten digits.\n\n  - http://yann.lecun.com/exdb/mnist/\n\n - To change the dataset, change the dataset loading code and sample size in the implementation files.\n\t\n - If you want to you can also import your own dataset, this code can be easily adapted to classify other type of images.\n\t\n```\nwidth = 16\nheight = 16\ndataset = semeion.read_data_semeion()\n```\n\n\u003cimg src=\"https://raw.githubusercontent.com/tentone/semeionNet/master/readme/dataset.png\" width=\"400\"\u003e\n\n\n\n\n\n### Install\n\n - The code available was tested with Python 3.5 and Tensorflow 1.1\n\n - Before running the examples in the repository, install the dependencies.\n\t\n```\ntensorflow matplotlib sklearn pandas numpy\n```\n\n​\t\n\n### Build and Run\n - Clone the repository into your computer\n\t- https://github.com/tentone/SemeionNet.git\n - Dataset files are already included in the repository inside the /source/dataset folder.\n - Run one of the implementation files from the source folder, each one implements a diferent classifier.\n    \t- knn.py, softmax.py, perceptron.py, cnn.py, lstm.py\n\n\n\n\n### Result Comparison\n - The results bellow were obtained, using 1300 random entries from the semeio dataset to train the classifier and 400 random entries to test the trained model.\n - The results obtained are expected, for the recurrent network (long short term memory), i haven't applied any confusion to the input, so after some time it detects that probably the next sample its equal to the current one.\n - Tests were run on a Core i5 6500 CPU with 24GB of RAM.\n\n| Classifier  | Time | Accuracy |\n| ----------- | ---- | -------- |\n| Softmax     | 40.8 | 94.32%   |\n| KNN         | 3.85 | 94.52%   |\n| Perceptron  | 11.4 | 97.16%   |\n| CNN         | 89.7 | 96.95%   |\n| RNN         | 35.1 | 97.56%   |\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Ftentone%2Fsemeionnet","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Ftentone%2Fsemeionnet","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Ftentone%2Fsemeionnet/lists"}