Example of Prior Work
Literature Selector
Using machine-learning techniques and training data from public and curated sources, create an automated pipeline to select literature of interest to the customer:
- Prepare a training set of 10K papers of interest
- Annotate and Preprocess Training Set
- Build Prediction Models and Prediction Workflows
- Narrow PUBMED papers pool down, using Machine Learning Tools (KNime; RBF classifier and DT methods; Scikit Learn SVC-RBF)
- Results narrow the PUBMED literature set down to about 10%