Predicting Water Quality Using Intel AI Analytics Toolkit

They will share how the team leveraged multiple components from the oneAPI AI Analytics Toolkit and API-based programming library to solve the problem of “predicting water quality” in the oneAPI Hackathon AI track. Leverage up to 6 models (lightgbm, xgboost, catboost, etc.) in their final solution and explore multiple optimization options to accelerate inference on Intel platforms, including using daal4py, oneDAL (C++). This complete solution includes data preprocessing, model classifier and fitting, standard model and D4P model conversion, and supports Python and C++ languages ​​for final inference. They also performed some preliminary performance evaluations of different optimization options and approaches, and shared some preliminary findings while investigating solutions to the challenges.

Download Presentation


Learn about joining the UXL Foundation:

Join now