https://github.com/shalabymhd/google-quickdraw-classification

A Kaggle Competition for the Applied Machine Learning (COMP 551) course at McGill University, under the supervision of Sarath Chandar.
https://github.com/shalabymhd/google-quickdraw-classification

Last synced: 22 days ago
JSON representation

A Kaggle Competition for the Applied Machine Learning (COMP 551) course at McGill University, under the supervision of Sarath Chandar.

• Host: GitHub
• URL: https://github.com/shalabymhd/google-quickdraw-classification
• Owner: shalabymhd
• Created: 2018-12-30T22:28:11.000Z (about 6 years ago)
• Default Branch: master
• Last Pushed: 2018-12-30T23:23:58.000Z (about 6 years ago)
• Last Synced: 2024-10-30T04:39:34.568Z (2 months ago)
• Language: C
• Homepage:
• Size: 47.2 MB
• Stars: 0
• Watchers: 1
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

README

        
# google-quickdraw-classification

The purpose of this project is to design a machine learning algorithm that can automatically identify hand drawn images as well as reason about their appearance, thus allowing automatic classification of the hand drawn images into one of the 31 classes. The hand drawn images are an extract from Googles Quick Draw Dataset. Performance is then compared to other participants in a Kaggle-based competition.

# Steps

  1) Feature Design

  2) Classifiers

  

    a) Logistic Regression

    b) Linear Support Vector Machine

    c) Feed-Forward Neural Network

    d) Convolutional Neural Network

    

  3) Hyper-parameter Tuning