I have a list of 100 cases with age and gender, and a list of 150 controls.
I want to match the cases to controls based on gender and age +-3 years.
Currently I'm using a greedy algorithm that tries to find first exact matches, then matches +-1year, +-2year, +-3years. However, this is not optimal way, and I could potentially get more matches with a non-greedy matching within the constrained +-3 years. I would prefer to optimize for maximum number of matches within +-3y and not optimize for minimum age difference.
Which algorithm/package can I use for this problem? (preferably in Python, but must be open-source)
For gender classification you may follow https://towardsdatascience.com/how-to-build-a-gender-classifier-in-python-using-scikit-learn-13c7bb502f2e
https://www.kaggle.com/shardulsane/gender-recognition-using-logistic-regression
Muhammad Ali Thanks Muhammad, but my question was not about gender classification, but about case-control matching, something entirely different.
You may use/follow "R" https://stackoverflow.com/questions/56026700/is-there-any-package-for-case-control-matching-individual-1n-matching-in-r-n
http://ugrad.stat.ubc.ca/R/library/e1071/html/matchControls.html
https://cran.r-project.org/web/packages/multipleNCC/multipleNCC.pdf
Hope this link will help your question
Article Age and sex matching in case-control studies
https://ncss-wpengine.netdna-ssl.com/wp-content/themes/ncss/pdf/Procedures/NCSS/Data_Matching-Optimal_and_Greedy.pdf
http://www.phusewiki.org/docs/Conference%202014%20HE%20Papers/HE04.pdf
- Badges
- Science topic
- Similar topics
- Computer Science and Engineering
- Programming Languages
- Python