GithubHelp home page GithubHelp logo

aldente0630 / loan-repayment-prediction Goto Github PK

View Code? Open in Web Editor NEW
0.0 1.0 0.0 165 KB

Automated approach from feature engineering to modeling on the Kaggle Home Credit Default Risk competition dataset

Jupyter Notebook 100.00%
featuretools mxnet autogluon

loan-repayment-prediction's Introduction

Automated Feature Engineering with FeatureTools

Relationship Diagram

  • A dataset with 2,221 features for 356,255 customers was finally created. Saved as a CSV file, it is about 4GB.
  • The whole process took 3 hours and a half on my iMac with 6 cores and 16GB of memory.

Automated Modeling with AutoGluon

  • The goal is to do automated modeling with AutoGluon.
  • AutoGluon makes it easy to automatically experiment with a variety of algorithms, from tree ensembles to deep learning and even model stacking.
model score_val pred_time_val fit_time
weighted_ensemble_k0_l2 0.787430 3098.737486 95757.520068
weighted_ensemble_k0_l1 0.786499 601.329862 46704.633752
CatboostClassifier_STACKER_l1 0.786261 2511.553999 53790.340020
LightGBMClassifierXT_STACKER_l1 0.785994 2511.152501 53834.121477
LightGBMClassifier_STACKER_l1 0.785990 2511.691034 53782.292310
LightGBMClassifierCustom_STACKER_l1 0.785596 2510.629085 54090.092252
LightGBMClassifierCustom_STACKER_l0 0.782958 10.546562 1941.627757
CatboostClassifier_STACKER_l0 0.782336 7.888541 1890.238214
LightGBMClassifierXT_STACKER_l0 0.780601 11.507542 860.345474
LightGBMClassifier_STACKER_l0 0.780356 10.297791 824.519218
...
  • 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.
  • 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.

Recommend Projects

  • React photo React

    A declarative, efficient, and flexible JavaScript library for building user interfaces.

  • Vue.js photo Vue.js

    ๐Ÿ–– Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.

  • Typescript photo Typescript

    TypeScript is a superset of JavaScript that compiles to clean JavaScript output.

  • TensorFlow photo TensorFlow

    An Open Source Machine Learning Framework for Everyone

  • Django photo Django

    The Web framework for perfectionists with deadlines.

  • D3 photo D3

    Bring data to life with SVG, Canvas and HTML. ๐Ÿ“Š๐Ÿ“ˆ๐ŸŽ‰

Recommend Topics

  • javascript

    JavaScript (JS) is a lightweight interpreted programming language with first-class functions.

  • web

    Some thing interesting about web. New door for the world.

  • server

    A server is a program made to process requests and deliver data to clients.

  • Machine learning

    Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.

  • Game

    Some thing interesting about game, make everyone happy.

Recommend Org

  • Facebook photo Facebook

    We are working to build community through open source technology. NB: members must have two-factor auth.

  • Microsoft photo Microsoft

    Open source projects and samples from Microsoft.

  • Google photo Google

    Google โค๏ธ Open Source for everyone.

  • D3 photo D3

    Data-Driven Documents codes.