ML Optimization

Practical constraints such as robustness, privacy, system power limits, quantization requirements, and fairness, make scalable training procedures challenging. Formalize, analyze, and develop scalable training procedures under practical constraints:

  • Accuracy-at-top constraints
  • Fairness constraints
  • Energy and quantization limitations
  • Dealing with robustness demands
  • Distributed/Federated learning and privacy constraints

img
img

ML Privacy

Practical constraints such as robustness, privacy, system power limits, quantization requirements, and fairness, make scalable training procedures challenging. Formalize, analyze, and develop scalable training procedures under practical constraints:

  • Accuracy-at-top constraints
  • Fairness constraints
  • Energy and quantization limitations
  • Dealing with robustness demands
  • Distributed/Federated learning and privacy constraints

ML Scalability

Practical constraints such as robustness, privacy, system power limits, quantization requirements, and fairness, make scalable training procedures challenging. Formalize, analyze, and develop scalable training procedures under practical constraints:

  • Accuracy-at-top constraints
  • Fairness constraints
  • Energy and quantization limitations
  • Dealing with robustness demands
  • Distributed/Federated learning and privacy constraints

img