/Don-t-Get-Kicked

:car: == :lemon: ? 1 : 0

Primary LanguagePython

Don't Get Kicked Kaggle Competition

🚗 == 🍋? 1 : 0

Link to competition: https://www.kaggle.com/c/DontGetKicked/

$sudo pip install -r requirements.txt
$python main.py

Approach

  1. Read 'train.csv' and 'test.csv' files
  2. Preprocess the files
  • Feature Engineering found in preprocess.py
  1. Perform Locality Sensitive Hashing (LSH) on the dataframe for test.csv
  • Finds 2000 similar rows for a query vector
  1. Perform K Nearest Neighbours on the LSH result
  • Finds k=1000 similar neighbours
  1. Pipeline of classification learning algorithms: Neural Network -> SVM