What is regret in online learning?

Regret measures cumulative loss compared to the best fixed strategy in hindsight. Sub-linear regret O(sqrt(T)) means the algorithm eventually matches the best expert.

What learning rate minimizes regret?

For OGD with L-Lipschitz losses and convex set diameter D, the optimal learning rate is D/(L*sqrt(T)), giving regret bound D*L*sqrt(T).

Online Learning Regret Bound

Compute cumulative regret bound for online gradient descent and follow-the-leader algorithms over T rounds given learning rate and loss structure.

Networking

Algorithms

Binary & Number

Systems

Dev Metrics

IP Subnet Calculator

IP Address

CIDR Prefix

Network

192.168.1.0

Broadcast

192.168.1.255

Subnet Mask

255.255.255.0

First Host

192.168.1.1

Last Host

192.168.1.254

Usable Hosts

254

Binary breakdown:

IP: 11000000.10101000.00000001.00000000

Mask: 11111111.11111111.11111111.00000000

Net: 11000000.10101000.00000001.00000000

How to Use

Enter the values for the patient or scenario you are assessing. Compute cumulative regret bound for online gradient descent and follow-the-leader algorithms over T rounds given learning rate and loss structure. Use the regret bound result to inform your clinical assessment.

Related Calculators

RL Sample Efficiency

Cybersecurity & Machine Learning

Q-Learning Convergence