Identifying conditions from human gait, using machine learning, but like humans

Viswadeep Sarangi Postgraduate researcher at the University of York, will be presenting his exciting work in using A.I to diagnose patients, just
by analysing their gait when walking.

This talk will touch upon the following overall topics.

1. What ML model tends to work on what sort of data, especially when it comes to human motion and gait
a. Non-biomimetic models such as SVM, DT, RDF
b. Biomimetic models such as MLP, RNN, LSTM
2. How to select or design the ML model to classify human gait
a. Data splitting
b. Data restructuring
c. Derivative data
3. How to leverage literature in human perception, neuroscience and psychology to maximize the ML model performance, for a task that humans are adept at, such as gait analysis
a. What we know about how humans do it
b. How to translate that knowledge into code for the ML models to use
c. What can this tell us about human neuroscience in return?
d. Advantages and limitations of this approach
4. Future directions in the research and commercialization plans undertaken at the moment
a. Reducing the need for data using Reinforcement and Imitation Learning
b. Synthetic data generation combined with classification models
c. Current post-doctoral work and release of alpha version of commercial software

To check out all my latest vacancies and to keep up to date with my market please follow the link: https://recruiters.thescoop.co.uk/cal...