{"id":17216911,"url":"https://github.com/lenguyenthedat/kaggle-for-fun","last_synced_at":"2025-10-08T12:44:27.077Z","repository":{"id":145642037,"uuid":"28439754","full_name":"lenguyenthedat/kaggle-for-fun","owner":"lenguyenthedat","description":"All my submissions for Kaggle contests that I have been, and going to be participating.","archived":false,"fork":false,"pushed_at":"2017-10-20T08:06:32.000Z","size":16485,"stargazers_count":40,"open_issues_count":4,"forks_count":37,"subscribers_count":3,"default_branch":"master","last_synced_at":"2025-03-27T12:52:44.365Z","etag":null,"topics":["kaggle","kaggle-contest","mnist","neural-network","python","scikit-learn","xgboost"],"latest_commit_sha":null,"homepage":"","language":"Python","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/lenguyenthedat.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":"2014-12-24T07:09:21.000Z","updated_at":"2024-11-26T07:22:05.000Z","dependencies_parsed_at":null,"dependency_job_id":"61b9d024-148c-42d8-b2a0-0da6631590c1","html_url":"https://github.com/lenguyenthedat/kaggle-for-fun","commit_stats":null,"previous_names":[],"tags_count":0,"template":false,"template_full_name":null,"repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/lenguyenthedat%2Fkaggle-for-fun","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/lenguyenthedat%2Fkaggle-for-fun/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/lenguyenthedat%2Fkaggle-for-fun/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/lenguyenthedat%2Fkaggle-for-fun/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/lenguyenthedat","download_url":"https://codeload.github.com/lenguyenthedat/kaggle-for-fun/tar.gz/refs/heads/master","host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":248790690,"owners_count":21162072,"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":["kaggle","kaggle-contest","mnist","neural-network","python","scikit-learn","xgboost"],"created_at":"2024-10-15T03:42:41.459Z","updated_at":"2025-10-08T12:44:22.052Z","avatar_url":"https://github.com/lenguyenthedat.png","language":"Python","funding_links":[],"categories":[],"sub_categories":[],"readme":"## Kaggle-for-fun\n\nhttps://www.kaggle.com/lenguyenthedat/competitions\n\nAll my submissions for Kaggle contests that I have been, and going to be participating.\n\nI will probably have everything written in Python (utilizing scikit-learn or similar libraries), but occasionally I might also use R or Haskell if I can.\n\nThe code written in this repository can be very scrappy since I consider this my playground. Feel free to contribute if you are interested in doing so.\n\n# avazu-ctr-prediction\nAvazu's CTR prediction contest - https://www.kaggle.com/c/avazu-ctr-prediction\n\nMy solution utilizes GradientBoostingClassifier with a few features preprocessing / engineering processes.\n\nIt scored *0.4045696* in term of log-loss, comparing to the first place at *0.3818529* (as of 2nd Feb 2014).\n\n# digit-recognizer\nDigit Recognizer - https://www.kaggle.com/c/digit-recognizer/\n\n- 1st version: Random Forest with 2000 estimators. Training time: 841s. Accuracy Score = 0.96800\n- 2nd version: 4 layer NN. Training time: 1774s. Accuracy Score = 0.97029\n- 3rd version: Convolution NN. Training time: 7674s. Accuracy Score = 0.97243\n- 4rd version: better Convo NN. Training time: 47903s. Accuracy Score = 0.98714\n\n# homesite-quote-conversion\nHomesite Quote Conversion - https://www.kaggle.com/c/homesite-quote-conversion/\n\nMy personal best single model, scoring 0.96896 on the public LB and 0.96832 on the private LB (Rank #37 out of 1762 teams)\n\n# how-much-did-it-rain-ii\nHow much did it rain II - https://www.kaggle.com/c/how-much-did-it-rain-ii\n\n- 1st version: basic xgboost. MAE: 23.81617\n- 2nd version: more features xgboost. MAE: 23.78276 # [1203]  eval-rmse:0.581621  train-rmse:0.407241\n- 3rd version: more features better tuning xgboost. MAE 23.77988 # [1772]   eval-rmse:0.553671  train-rmse:0.346084\n\n# liberty-mutual-group-property-inspection-prediction\nLiberty Mutual Group: Property Inspection Prediction - https://www.kaggle.com/c/liberty-mutual-group-property-inspection-prediction\n\n- 1st version LB 0.336953: basic RandomForest.\n- 2nd version LB 0.366923: xgboost.\n- 3rd version LB 0.372686: better xgboost parameter.\n- 4rd version LB 0.378719: native xgboost. (Local score: 0.376601296742)\n- 5th version LB 0.380626: tuned native xgboost. (Local score: 0.383170885213)\n- 6th version LB 0.380916: tuned native xgboost. (Local score: 0.380586744068) (5th seems to be overfit...)\n\n# otto-group-classification\nOtto Group Product Classification - https://www.kaggle.com/c/otto-group-product-classification-challenge\n\n- 1st version: basic Adaboost. 28s training time, log_loss = 0.95514\n- 2nd version: Neural Network. 96s training time, log_loss = 0.53893\n- 3rd version: 4 layers Neural Network, 335s training time, log_loss = 0.50003\n\n# rossmann-store-sales\nRossmann Store Sales - https://www.kaggle.com/c/rossmann-store-sales\n\n- 1st version LB 0.37205: basic xgboost\n- 2nd version LB 0.11764 (local: 0.1169): fine-tuning xgboost 1000s training time\n- 3nd version LB 0.11021 (local: 0.1097): like 2nd version but with log and exp mod\n- 4rd version LB 0.10903 (local: 0.1116): xgboost 3000r d10 lr 0.02 sample 0.9 0.7 - 3600 training time\n- 5rd version LB 0.10861 (local: 0.1072): xgboost 3000r d12 lr 0.02 sample 0.9 0.7 - 4500 training time\n- 6th version LB 0.10795 (local: 0.1107): native-xgb 3000r d10 lr 0.02 sample 0.9 0.7\n- 7th version LB 0.10640 (local: 0.1070): r version xgb 3000r d10 lr 0.02 sample 0.9 0.7\n- 8th version LB 0.10568 (local: 0.0993): r version xgb 8000r d13 lr 0.01 sample 0.9 0.7\n- 9th version LB 0.10858 my own xgb native implementation d13 eta0.01 ntree4000 mcw3.\n\n# sf-crime-classification\nSan Francisco Crime Classification - https://www.kaggle.com/c/sf-crime/\n\n- 1st version: basic Adaboost. 47s training time, log_loss = 3.66252\n- 2nd version: fine tuned Adaboost. 525s training time, log_loss = 2.72599\n- 3rd version: 2-layers Neural Network. 770s training time, log_loss = 2.51524\n- 4th version: 2-layers Neural Network with Engineered features. 823s training time, log_loss = 2.47535\n- 5th version: 4-layers Neural Network with Engineered features. 3006s training time, log_loss = 2.43479\n- 6th version: Fine-tuned Random Forest (1024 trees, depth 16). 3560s training time, log_loss = 2.33752\n- 7th version: xgb native. 3000s training time, log_loss = 2.33537\n\n# springleaf-marketing-response\nSpringleaf Marketing Response - https://www.kaggle.com/c/springleaf-marketing-response\n\n- 1st version: native xgboost. AUC Score = 0.78714 Booster-submit-d8-0.01-mcw3-0.65-0.65-500r\n- 2nd version: feature selection and native xgboost. AUC Score = 0.79164. Local: 0.756932926155\n- 3rd version: feature selection with ftrl. AUC Score = 0.77348\n- 4th version: native xgboost. AUC Score = 0.79200. Local: 0.754258680372\n\n# titanic\nTitanic - https://www.kaggle.com/c/titanic/\n\n- 1st version: basic 4-layer Neural Network. Accuracy Score = 0.80383\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Flenguyenthedat%2Fkaggle-for-fun","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Flenguyenthedat%2Fkaggle-for-fun","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Flenguyenthedat%2Fkaggle-for-fun/lists"}