{"id":20093221,"url":"https://github.com/wilberquito/mnist-autoencoder-classification","last_synced_at":"2025-07-08T15:04:45.736Z","repository":{"id":237313321,"uuid":"794269167","full_name":"wilberquito/mnist-autoencoder-classification","owner":"wilberquito","description":"Autoencoder Feature Extraction for Classification. SSL","archived":false,"fork":false,"pushed_at":"2024-05-01T21:16:06.000Z","size":2102,"stargazers_count":0,"open_issues_count":0,"forks_count":0,"subscribers_count":1,"default_branch":"main","last_synced_at":"2025-03-02T16:13:30.547Z","etag":null,"topics":[],"latest_commit_sha":null,"homepage":"","language":"Jupyter Notebook","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":null,"status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/wilberquito.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":null,"funding":null,"license":null,"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":"2024-04-30T19:32:59.000Z","updated_at":"2024-05-01T21:16:10.000Z","dependencies_parsed_at":"2024-05-01T22:44:00.369Z","dependency_job_id":"ec2759fb-5cdf-4ac8-b67e-ee6a0184d20c","html_url":"https://github.com/wilberquito/mnist-autoencoder-classification","commit_stats":null,"previous_names":["wilberquito/mnist-autoencoder-classification"],"tags_count":0,"template":false,"template_full_name":null,"purl":"pkg:github/wilberquito/mnist-autoencoder-classification","repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/wilberquito%2Fmnist-autoencoder-classification","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/wilberquito%2Fmnist-autoencoder-classification/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/wilberquito%2Fmnist-autoencoder-classification/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/wilberquito%2Fmnist-autoencoder-classification/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/wilberquito","download_url":"https://codeload.github.com/wilberquito/mnist-autoencoder-classification/tar.gz/refs/heads/main","sbom_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/wilberquito%2Fmnist-autoencoder-classification/sbom","host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":264292913,"owners_count":23586060,"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":[],"created_at":"2024-11-13T16:46:16.644Z","updated_at":"2025-07-08T15:04:45.713Z","avatar_url":"https://github.com/wilberquito.png","language":"Jupyter Notebook","funding_links":[],"categories":[],"sub_categories":[],"readme":"# Autoencoder Feature Extraction for Classification\n\nAutoencoders are a type of neural network which generates an “n-layer” coding\nof the given input and attempts to reconstruct the input using the code\ngenerated.\n\nThe Autoencoder architecture architecture is divided into the encoder\nstructure, the decoder structure, and the latent space, also known as the\n“bottleneck”.\n\n## Encoder\n\n$$h = E(x)$$\n\n## Decoder\n\n$$x' = D(h)$$\n\n## Latence space\n\nThis is the data representation or the low-level, compressed representation of\nthe model’s input. The decoder structure uses this low-dimensional form of data\nto reconstruct the input. It is represented by $h$.\n\n## Self Supervised Learning\n\n### pretext task\n\nNow for both model to learn, we need a metric. This metric of loss $L$ which\nshould mesure how good the decoder $D$ does reconstructing the original data\nfrom the encoder $E$.\n\n$$L = Loss(x, x')$$\n\n### downstream task\n\nThe learned representation by encoder $E$ can be used and fine tunned. As\nencoder $E$ \"knows\" important features from the SSL problem, we can use it for\ntransfer learning in a classification or regression task.\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fwilberquito%2Fmnist-autoencoder-classification","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fwilberquito%2Fmnist-autoencoder-classification","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fwilberquito%2Fmnist-autoencoder-classification/lists"}