aauss / epi-quark Goto Github PK
View Code? Open in Web Editor NEWBased on labeled data, multi-class and multi-label scoring of outbreak detection, annotation or prediction algorithms.
License: MIT License
Based on labeled data, multi-class and multi-label scoring of outbreak detection, annotation or prediction algorithms.
License: MIT License
current version:
"Scoring the detection, labeling, and prediction performance of disease outbreak detection algorithms across different algorithm families."
proposal:
"Based on labeled data, multi-class and multi-label scoring of outbreak detection, annotation or prediction algorithms."
This should be coherent with documentation (https://github.com/aauss/epi-quark/blob/master/README.md, https://github.com/aauss/epi-quark/blob/master/docs/source/index.rst).
This could be implemented by adding an optional column "weight" to the input case data:
If the column is absent, and weights are not computed via epi-quark, then all weights are 1.
Accordingly, one more value for parameter weighting
of calc_score()
should be added to the values allowed. (This seems safer to me than implicitly taking the user-defined weights when they are provided... this also ensures that the user chooses which type of weight they want, e.g., not user defined and time-space at the same time.)
They should be stored in a new input table of the following form:
data_label | signal_label | value |
---|---|---|
... | ... | ... |
The resulting data frame is used as argument of score functions computing the probabilities.
For outbreak labels not present: flat beside endemic and non-case.
Relevant for demographics or administrative units, for example.
In particular output, for chosen score types (e.g., accuracy) or for all available score types, the respective groups with minimum score. This allows for Rawlsian maximin strategy, see e.g. https://arxiv.org/abs/2006.13114
Relevant for scoring is the expected case number for each outbreak class and data cell.
Since each case should be allowed different probabilities of belonging to different outbreak classes, the expected case number need not be integer.
N.B. Since a case can belong to many classes, these probabilities need not sum up to 1 and there is no constraint on the expected case number beside being >= 0.
If I'm not mistaken no code change is necessary, only the description of value
in the API and scorer documentations should be modified and sentence This value must be a positive integer.
removed across https://github.com/aauss/epi-quark/blob/master/epiquark/api.py and https://github.com/aauss/epi-quark/blob/master/epiquark/scorer.py
A declarative, efficient, and flexible JavaScript library for building user interfaces.
๐ Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
An Open Source Machine Learning Framework for Everyone
The Web framework for perfectionists with deadlines.
A PHP framework for web artisans
Bring data to life with SVG, Canvas and HTML. ๐๐๐
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
Some thing interesting about web. New door for the world.
A server is a program made to process requests and deliver data to clients.
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
Some thing interesting about visualization, use data art
Some thing interesting about game, make everyone happy.
We are working to build community through open source technology. NB: members must have two-factor auth.
Open source projects and samples from Microsoft.
Google โค๏ธ Open Source for everyone.
Alibaba Open Source for everyone
Data-Driven Documents codes.
China tencent open source team.