{"id":26372715,"url":"https://github.com/bits-bytes-nn/loan-repayment-prediction","last_synced_at":"2025-03-17T01:18:58.609Z","repository":{"id":201279658,"uuid":"341433550","full_name":"bits-bytes-nn/loan-repayment-prediction","owner":"bits-bytes-nn","description":"Automated approach from feature engineering to modeling on the Kaggle Home Credit Default Risk competition dataset","archived":false,"fork":false,"pushed_at":"2021-03-03T03:01:46.000Z","size":165,"stargazers_count":0,"open_issues_count":0,"forks_count":0,"subscribers_count":1,"default_branch":"main","last_synced_at":"2025-01-19T12:28:20.774Z","etag":null,"topics":["autogluon","featuretools","mxnet"],"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/bits-bytes-nn.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}},"created_at":"2021-02-23T04:59:02.000Z","updated_at":"2021-03-15T12:19:13.000Z","dependencies_parsed_at":null,"dependency_job_id":"5ed2cded-a8b4-4760-8f24-e99219c7b988","html_url":"https://github.com/bits-bytes-nn/loan-repayment-prediction","commit_stats":null,"previous_names":["aldente0630/loan-repayment-prediction","bits-bytes-nn/loan-repayment-prediction"],"tags_count":null,"template":false,"template_full_name":null,"repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/bits-bytes-nn%2Floan-repayment-prediction","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/bits-bytes-nn%2Floan-repayment-prediction/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/bits-bytes-nn%2Floan-repayment-prediction/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/bits-bytes-nn%2Floan-repayment-prediction/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/bits-bytes-nn","download_url":"https://codeload.github.com/bits-bytes-nn/loan-repayment-prediction/tar.gz/refs/heads/main","host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":243955782,"owners_count":20374373,"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":["autogluon","featuretools","mxnet"],"created_at":"2025-03-17T01:18:58.137Z","updated_at":"2025-03-17T01:18:58.587Z","avatar_url":"https://github.com/bits-bytes-nn.png","language":"Jupyter Notebook","funding_links":[],"categories":[],"sub_categories":[],"readme":"# Automated Feature Engineering with *FeatureTools*\n* The goal is to do automated feature engineering with [FeatureTools](https://www.featuretools.com).\n* A dataset of Kaggle Competition, [Home Credit Default Risk](https://www.kaggle.com/c/home-credit-default-risk/overview/description) was downloaded for testing. The dataset consists of 4 tables, and the relationship diagram is as follows. Various derived variables could be created automatically using *FeatureTools*.\n* I referenced [Will Koehrsen's post](https://towardsdatascience.com/automated-feature-engineering-in-python-99baf11cc219 ) and [his code on Kaggle](https://www.kaggle.com/willkoehrsen/feature-engineering-using-feature-tools).\n  \n![Relationship Diagram](https://aldente0630.github.io/assets/home_credit.png)  \n\n* A dataset with 2,221 features for 356,255 customers was finally created. Saved as a CSV file, it is about 4GB.\n* The whole process took 3 hours and a half on my iMac with 6 cores and 16GB of memory.\n\n# Automated Modeling with *AutoGluon*\n* The goal is to do automated modeling with [AutoGluon](https://auto.gluon.ai/stable/index.html#). \n* *AutoGluon* makes it easy to automatically experiment with a variety of algorithms, from tree ensembles to deep learning and even model stacking.\n  \n|model|score_val|pred_time_val|fit_time|\n|:------:|------:|------:|------:|\n|weighted_ensemble_k0_l2|0.787430|3098.737486|95757.520068|\n|weighted_ensemble_k0_l1|0.786499|601.329862|46704.633752|\n|CatboostClassifier_STACKER_l1|0.786261|2511.553999|53790.340020|\n|LightGBMClassifierXT_STACKER_l1|0.785994|2511.152501|53834.121477|\n|LightGBMClassifier_STACKER_l1|0.785990|2511.691034|53782.292310|\n|LightGBMClassifierCustom_STACKER_l1|0.785596|2510.629085|54090.092252|\n|LightGBMClassifierCustom_STACKER_l0|0.782958|10.546562|1941.627757|\n|CatboostClassifier_STACKER_l0|0.782336|7.888541|1890.238214|\n|LightGBMClassifierXT_STACKER_l0|0.780601|11.507542|860.345474|\n|LightGBMClassifier_STACKER_l0|0.780356|10.297791|824.519218|\n| ... | | | |\n  \n* The model stacking technique achieved the highest predictive performance. This was 0.78149 for the Kaggle public board and 0.78391 for the private board as measured by AUROC.\n* This process took about 1 day and 6 hours to train on an AWS `m4.16xlarge` EC2 instance with 64 cores and 256GB of memory, and about an hour and a half to infer.\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fbits-bytes-nn%2Floan-repayment-prediction","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fbits-bytes-nn%2Floan-repayment-prediction","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fbits-bytes-nn%2Floan-repayment-prediction/lists"}