I recently started working on the age detection hackathon
i am converting all images to 3232 size, greyscale and then using the classif.ksvm algorithm via the mlr package to classify the test data.
i used 3232 size as 64*64 makes the dataset too big for R to work on
i used mlr package as it has given good result in the past…
as yet, my score is very low, and it seems that there is a lot of area for improvement.