Interpretable meta-analysis of model or marker performance https://arxiv.org/abs/2409.13458 arXiv:2409.13458v1 Announce Type: new Abstract: Conventional meta analysis of model performance conducted using datasources from different underlying populations often result in estimates that cannot be i 📈🤖
Okay, I have the platform/enterprise/machine datasources slides, DFIR tooling, Autopsy walkthru, and common Azure misconfigs slides (which are very boring and need an uplift). The incident recap and the report writing slides can wait til tomorrow. I am tired and fed up with working the weekend.
I'm actually surprised this doesn't exist; a local-first dashboard/BI tool. Datasources are fetched via API and cached locally. Wasm sqlite/postgres would be perfect for this. Optional cloud backup Someone please build this?
It has all the hallmarks of a data product that is of no use: -convoluted process -hand-entered datasources -inscrutable KPI definitions -no versioning, no documentation All of the processes she manages are like this. This dashboard has existed since 2018 and has only been viewed 170-ish times.
It's wild how badly my manager fumbles the handoff of literally every project. This latest one is mysteriously missing datasources and nobody can tell me how long this thing has been broken because - as I suspected - nobody actually uses this.
just thought of a way to avoid the "catch-up spikes" in bskycharts while letting legit spikes through: track the lag time and if more than 30s behind return N/A for all the datasources being queried
There's a new git repo for tutorials, scripts, and other datasources that can be loaded to the Bus Pirate's internal flash. If you make anything fun, please consider making a push request to include it in the repo: github.com/DangerousPro...
Command macros, tutorials, device libraries and other data sources for Bus Pirate 5 - DangerousPrototypes/BusPirate5-database