Built a Animal Wildlife Estimator using Social Media

it_user489873 - PeerSpot reviewer

Project Description

  • • Formulated a population estimation model for animals in certain geographic regions using pictures shared on various social media platforms.
  • • Designed an environment for simulating the sharing of pictures on social media on Amazon Mechanical Turk using API’s built on Python for interfacing with AWS for creating jobs and parsing results of the turk.
  • • Implemented and evaluated different machine learning algorithms in Python3 to predict the likelihood of a picture being shared on social media based on various ecological as well biological features.
  • • Interpreted, quantified and incorporated different levels of self-reporting or behavioral bias towards sharing photographs of species of animals with unique patterns in the proposed population estimation model.

Highlights

Received recognition / award
Support from colleagues

Difficulties

Steep learning curve
Hard to meet schedule

Products Used

Technical Skills Used

  • Python
  • Machine Learning
  • Statistics
  • Bash Scripting
  • Chicago (IL-US)41.85-87.65