{"id":31781322,"url":"https://github.com/wcchu/fvn-fbk","last_synced_at":"2025-10-10T08:50:40.787Z","repository":{"id":44887463,"uuid":"58491567","full_name":"wcchu/FVN-fbk","owner":"wcchu","description":"Fixed-volume neighborhood classifier with binary feedback","archived":false,"fork":false,"pushed_at":"2022-07-12T10:55:05.000Z","size":10,"stargazers_count":0,"open_issues_count":0,"forks_count":0,"subscribers_count":3,"default_branch":"main","last_synced_at":"2023-03-08T22:10:59.968Z","etag":null,"topics":["classification","fnn","knn","r","recommender-system"],"latest_commit_sha":null,"homepage":"","language":"R","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/wcchu.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}},"created_at":"2016-05-10T20:30:12.000Z","updated_at":"2018-12-01T12:20:10.000Z","dependencies_parsed_at":"2022-09-02T10:22:04.591Z","dependency_job_id":null,"html_url":"https://github.com/wcchu/FVN-fbk","commit_stats":null,"previous_names":[],"tags_count":null,"template":null,"template_full_name":null,"purl":"pkg:github/wcchu/FVN-fbk","repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/wcchu%2FFVN-fbk","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/wcchu%2FFVN-fbk/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/wcchu%2FFVN-fbk/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/wcchu%2FFVN-fbk/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/wcchu","download_url":"https://codeload.github.com/wcchu/FVN-fbk/tar.gz/refs/heads/main","sbom_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/wcchu%2FFVN-fbk/sbom","scorecard":null,"host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":279003263,"owners_count":26083555,"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","status":"online","status_checked_at":"2025-10-10T02:00:06.843Z","response_time":62,"last_error":null,"robots_txt_status":"success","robots_txt_updated_at":"2025-07-24T06:49:26.215Z","robots_txt_url":"https://github.com/robots.txt","online":true,"can_crawl_api":true,"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":["classification","fnn","knn","r","recommender-system"],"created_at":"2025-10-10T08:50:35.417Z","updated_at":"2025-10-10T08:50:40.781Z","avatar_url":"https://github.com/wcchu.png","language":"R","funding_links":[],"categories":[],"sub_categories":[],"readme":"# Fixed-volume neighborhood classifier with binary feedback\n\nNature of the problem:\nSuppose we have n numerical (predictor) variables Vp = [V1, ... Vn], 1 categorial variable Vc, and a binary response R. When given a query with Vp, which category in Vc do we choose to \"optimize\" the response?\n\nExample 1:\nConsider the record of customers watching movies in a theater. We know the basic customer info: age, distance from theater, monthly movie budget etc; we know the environmental data: outdoor temperature, economic index etc; we know the movie genre; and we know the feedback after the movie to be positive or negative. For a given set of customer and environmental data Vp, what genre Vc do we recommend to get a positive feedback (R = 1)?\n\n(1) Fixed-volume-neighborhood approach\n\nAll predictor variables Vp are numeric so the distance-based algorithm is valid. While kNN finds k nearest neighbors, we instead find all the data points within a fixed-volume in the variable space, so that the highly sparse areas are without recommendation.\n\nAbout the \"positive feedback\":\nWhile the goal is to let the feedback be as positive as possible, there are 2 different sub approaches--1. choose the genre that will give the highest positive-feedback-rate, 2. choose the genre that will give as many positive-feedback customers as possible.\n\n(2) Binary probablity approach\n\nFirst treat Vc as another predictor variable to train the data, and get a model M where (Vp, Vc) is the input and R is the output. When we are given a set of \"actual\" predictor variables Vp(0), we run through all possible Vc and throw each (Vp(0), Vc(i)) for i = 1:number_of_classes to model M and get R(i). Then we pick the Vc(i) that gives the highest probability of R(i) = 1. The model M can be trained with any binary classifier.\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fwcchu%2Ffvn-fbk","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fwcchu%2Ffvn-fbk","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fwcchu%2Ffvn-fbk/lists"}