{"id":23324889,"url":"https://github.com/darraghdog/avito-demand","last_synced_at":"2026-03-06T22:03:30.373Z","repository":{"id":80355813,"uuid":"133146978","full_name":"darraghdog/avito-demand","owner":"darraghdog","description":"Avito Demand Prediction Challenge :department_store:","archived":false,"fork":false,"pushed_at":"2018-06-30T22:40:14.000Z","size":97785,"stargazers_count":56,"open_issues_count":0,"forks_count":18,"subscribers_count":4,"default_branch":"master","last_synced_at":"2025-08-22T19:10:23.799Z","etag":null,"topics":["datatable","lightgbm","mlp","python3","r","rnn","sklearn"],"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/darraghdog.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":"2018-05-12T13:07:12.000Z","updated_at":"2025-04-15T00:15:41.000Z","dependencies_parsed_at":"2023-06-06T05:45:14.126Z","dependency_job_id":null,"html_url":"https://github.com/darraghdog/avito-demand","commit_stats":null,"previous_names":[],"tags_count":0,"template":false,"template_full_name":null,"purl":"pkg:github/darraghdog/avito-demand","repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/darraghdog%2Favito-demand","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/darraghdog%2Favito-demand/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/darraghdog%2Favito-demand/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/darraghdog%2Favito-demand/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/darraghdog","download_url":"https://codeload.github.com/darraghdog/avito-demand/tar.gz/refs/heads/master","sbom_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/darraghdog%2Favito-demand/sbom","scorecard":null,"host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":286080680,"owners_count":30200756,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2026-03-06T19:07:06.838Z","status":"ssl_error","status_checked_at":"2026-03-06T18:57:34.882Z","response_time":250,"last_error":"SSL_connect returned=1 errno=0 peeraddr=140.82.121.6:443 state=error: unexpected eof while reading","robots_txt_status":"success","robots_txt_updated_at":"2025-07-24T06:49:26.215Z","robots_txt_url":"https://github.com/robots.txt","online":false,"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":["datatable","lightgbm","mlp","python3","r","rnn","sklearn"],"created_at":"2024-12-20T18:28:05.865Z","updated_at":"2026-03-06T22:03:30.354Z","avatar_url":"https://github.com/darraghdog.png","language":"Jupyter Notebook","funding_links":[],"categories":[],"sub_categories":[],"readme":"### Kaggle - Avito Demand Prediction\n  \n**5th Place solution - team `Optumize`**  \nPredict demand for an online classified ads  \nhttps://www.kaggle.com/c/avito-demand-prediction  \n  \n\n## Models Progrssion\n### Single Models\n```\nModel         Small Val          5CV Val            Leaderboard     Comment\n================================================================================================\nlgb_1406A     0.2113             0.2136             ??????          tuning - from 250 to 1000 leaves\nlgb_1406      0.2113             0.2136             ??????          tuning - from 250 to 1000 leaves\nlgb_1106A     0.2117             0.2140             ??????          Add more encoding features\nlgb_1006      0.2120             0.2153             ??????          Add param features                    \nlgb_0906      0.2123             0.2158             ??????          Add price ranking reatures\nlgb_0206      0.2132             0.2162             ??????          Add title translation as col and tfidf it\nlgb_3105      0.2134             0.2168             0.2190          Meta image features\nrnndh_0406a   0.2136                                0.2194          Add price ratios\nlgb_2705B     0.2137             0.2168             0.2194          imgtop1 ratios;longer early stopping;remove categorical\nlgb_2705A     0.2139             ??????             0.2197          Remove categoricals, add image_top_1 price ratio\nlgb_2505      0.2143             0.2167             0.2202          More FE - Price \u0026 Item Seq ratios over category/title\nlgb_2405D     0.2145             ??????             0.2204          Price ratios over category/title   \nlgb_2405      0.2152             ??????             0.2211          pymorph on text\nrnn_2605      0.2146             ??????             0.2213          Logit averaging and stopword removal\nlgb_2205      0.2153             ??????             0.2213          Add oof ridge feature on text data and image data.\nrnn_2205      0.2149             ??????             0.2215          Russian text processing\nlgb_2205      0.2157             ??????             0.2215          Add oof ridge feature on text data\nmlp_1905      0.2159             ??????             0.2217          Add in different kinds for grouping continuous\nmlp_1705      0.2162             0.21875            0.2217          Add in aggregate features from active files\nrnn_2105      0.2153             ??????             0.2221          RNN only submission, more regularization .2153 on validation\nmlp_1605B     0.2166             ??????             0.2224          Add all item titles from avctive files per user\nlgb_2105C     0.2162             ??????             0.2225          Add count and encoding\nmlp_1605A     0.2170             ??????             0.2228\nnnet_1505     0.2177             ??????             \nlgb_2105      0.2174             ??????             0.2133             \nlgb_1404      0.2182             ??????             0.2241\n```\n\n### Blend (Weighted Average)\n```\nModel         Small Val          5CV Val            Leaderboard     Comment\n================================================================================================\nblend3x_2605  ????               ?????              0.2188          Blend of 0.25 * mlp_1705, 0.5 * lgb_2505 and 0.25 * rnn_2205\nall_2405      ????               ?????              0.2193          Equal Blend of mlp_1905, lgb_2205 and rnn_2205\nmlp_1905      ????               ?????              0.2204          MLP 1705A and 1905 50/50 and mix 50/50 with best LB            \nmlp_1705A     ????               ??????             0.2204          Weighted avg mlp and best lb kernel https://www.kaggle.com/lscoelho/blending-\nmlp_1605B     ????               ??????             0.2208          Weighted avg mlp and best lb kernel https://www.kaggle.com/lscoelho/blending-models-lb-0-2216   \n```\n\n### Stack\n```\nModel         Small CV Val       5CV Val         Leaderboard     Comment\n              in blend script\n================================================================================================\nL1GBM_2306    0.2113             ?????              0.2147          More ridge and add user entropy features\nL1GBM_2006A   0.2118             ?????              0.2151          different tfidf\nL1GBM_1606B   0.2122             ?????              0.2153          More features at L2\nL1GBM_1506B   0.2126             ?????              0.2154          Bag the L2 lgb; Bag the 1406 sub\nL1GBM_1506    0.2127             ?????              0.2155          Add lgb 1406 leaves tuning\nL1GBM_1006A   0.2133             ?????              0.2161          Add param features\nL1GBM_1006    ??????             ?????              0.2163          Add price ranking\nL1GBM_0406A   ??????             ?????              0.2166          fixed lgb bug at L1, hash 'text' instead of description\nL1GBM_0306A   ??????             ?????              0.2167          Included title translation LGB \n\n```\n\n### Word embeddings\n```\n=================================================================================================\nfeatures/wiki.ru.vec                  - https://github.com/facebookresearch/fastText/blob/master/pretrained-vectors.md\ncc.ru.300.vec.gz                      - https://github.com/facebookresearch/fastText/blob/master/docs/crawl-vectors.md\nall.norm-sz100-w10-cb0-it1-min100.w2v - http://panchenko.me/data/dsl-backup/w2v-ru/\n```\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fdarraghdog%2Favito-demand","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fdarraghdog%2Favito-demand","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fdarraghdog%2Favito-demand/lists"}