Notes
Notes on learning across institutions that cannot share their data, survival analysis, and predicting when something will happen. By Navid Seidi.
-
My Ph.D. in three minutes
The dissertation is 278 pages. Here is the short version, a three-minute film in my own voice about learning together without sharing data and predicting when, not just whether.
-
HARMONY: reputation for federated survival analysis
Our IEEE JBHI paper HARMONY lets hospitals in a federation judge each other's contributions through a moving reputation score, without ever seeing a peer's real update.
-
Back at CHASE, with a harder question
Back at IEEE/ACM CHASE, this time with a sparse peer-to-peer network for federated survival prediction where every connection has to keep earning its place.
-
Agents that cannot assume somebody else pays for the compute
PerAgents 2027, co-located with IEEE PerCom, is about agents that run without a large model behind an API, spare context or a network that is always there. The call for papers is open.
-
NSF I-Corps National Team badge
My verified achievement from the NSF National I-Corps program, issued by the National Science Foundation.
-
What NSF I-Corps taught our team
A short note on what the NSF I-Corps program gave our team about bringing SmarTreatMate closer to the market.
-
Calling healthcare providers and cancer patients
An open call for oncologists, cancer patients, survivors and supporters to be interviewed for an NSF-funded project on cancer care.
-
SmarTreatMate launches its website
The SmarTreatMate team, part of the NSF National I-Corps Summer Cohort #3, launched its website for the AI-driven treatment outcome tool.
-
Completing the W23 Medical Jumpstart I-Corps course
SmarTreatMate completed the W23 Medical Jumpstart I-Corps course with the NSF Great Lakes Regional I-Corps Hub.