I need to speed up my Naive Bayes classiffication in R. I found an article which describes how to use Intel DAAL's Naive Bayes classifier from R and since I have a Xeon Phi that the library could use I thought that's a good approach: https://www.codeproject.com/Articles/1151606/Lightning-Fast-R-Machine-Learning-Algorithms but the problem is that DAAL is the old name and version. It now became oneDAL (part of oneAPI). I wasn't able to use the information from the article to get it working with oneDAL and I can't find where I can get DAAL (the older version, to which the article applies).
I would either need instructions for oneDAL to integrate it with R (equivalent to the ones in the article, for DAAL) or a way to obtain DAAL, the old version.
The last resort, I imagine I could use Python (for which there is more info available for oneDAL than R) to do the classification and then transfer the results to R to continue work there.
Although there was indeed name change from DAAL to oneDAL, but all API's and internals remain the same. So I would expect things to keep working based on what is written in instruction. You can get 2020.3 version that was last before the renaming - https://anaconda.org/intel/daal/files
There is python api for Naive Bayes https://intelpython.github.io/daal4py/algorithms.html#multinomial-naive-bayes