Kaggle

My Profile

https://www.kaggle.com/bangdasun

Competitions

Classification

• Titanic passenger survival prediction

  #binary, #get started

• Instacart market basket analysis

  #binary, #recommendation

• Toxic comment classification challenge

  #multi-labels, #NLP

• TalkingData adtracking fraud detection challenge

  #binary, #CTR

• Home Credit default risk

  #binary, #finance fraud detection

• Quora insincere questions classification

  #binary, #NLP

Regression

• House price prediction - advanced regression techniques

  #house price, #get started

• New York City taxi trip duration

  #trip duration, #get started

• Mercari price suggestion challenge

  #price, #e-commerce, #NLP, #recommendation

• Avito demand prediction challenge

  #deal probability, #e-commerce, #NLP, #CV, #CTR

• Santander value prediction challenge

  #anonymous data, #finance

• Elo merchant category recommendation

  #loyalty score, #banking, #business

Image

• TGS salt identification challenge

  #segmentation

• Airbus ship detection challenge

  #segmentation, #object detection

Resource

General Information

• Kaggle ML and Data Science Survery 2017

My collections

• ml-workflow-example_01: general framework of how to build a workflow use sklearn (use Random Forest for iris)
• ml-workflow-example_02: XGBoost usage in Titanic (build model, train model, hyperparameter tuning)
• ml-workflow-example_03: model evaluation (hold-out set validation, cross validation, metrics include roc-auc, recall, precision) on UCI data
• ml-workflow-example_04: cross validation for a given model

My competition write-up and top solutions reading

• My zhihu articles