Hamdi will complete this project during his internship ith us in summer 2018, details from his proposal:
Implement a feature selection suite for ECL-ML. In the first part, we will implement several supervised and unsupervised filter-based feature selection algorithms, and test them using HPCC platform and UCI-based datasets. Finally, we will work on the design and implementation of a new filter-based feature selection algorithm, which selects the features using both supervised and unsupervised technique.
- Implementation of all supervised and unsupervised filter-based methods (see proposal)
- Design and implementation of a new filter-based approach, which leverages from both supervised and unsupervised approach.
- Wishlist (if time permits) - Implementation of wrapper and embedding methods