Machine learning based system for dog behaviour prediction

  • Completed on: 2016-02-28
  • Technologies: Python / Sklearn / Keras
  • Client: Tail

Project Info:

Tail aims to help dog owners take better care of their dogs. The dogs were equipped with collars that identified their location and provided accelerometers’ data. Our involvement consisted of developing a machine learning system capable of telling what activities a dog is engaged in based on the data coming in from the collar. Devising the system involved analyzing the accelerometers’ data and training recurrent neural networks capable of reducing the data into actionable information.