The complexity of the problem I dive in to start to scare me. Every paper I read, written by a team of researches under cover of some recognisable university. Numbers of hypothesis to check increases exponentially. Here some of them:

  • applicability of ballistocardiography to measure horse (animal) heart \ breath rate
  • transferability activity recognition methods based on ML model between different horses
  • anomaly detection of horse behaviour
  • applicability of stream or batch clusterization methods for unsupervised learning to detect animal activity
  • using transfer learning methods to detect illness symptoms

So, here is a lot of work for a joined team of data scientists, ML scientists, veterinarian, programmers… And a good problem to try to get research fundings….. Even more. here is a good perspective for the monetization of research results. Does anyone have any suggestion how to try to organise such kind research lab in collaboration with some university or research centre?