Many technologies such as Deep Learning and tools like Word Embeddings havestarted to be investigated only recently . Many challenges remain open whenit comes to healthcare domain applications . To address these challenges, we propose the use of Deep Learning to identify sixteen morbidity types within textual descriptions of clinical records . From the obtained results it seems that the latter outperforms the combination of . Deep Learning approaches using Word embeddings. Using Support VectorMachine and Multilayer perceptron (our baselines) seems to be the best approach to identify 16 disease types in clinical records. Our preliminary results indicate that there arespecific features that make the dataset biased in favour of traditional machinelearning approaches. The results indicate there are . there are specific features that . make the . dataset biased . of the dataset that make . the dataset bias in favour . of Traditional machine learning approaches. It appears that there is . the . data are biased in against the traditional tf-idf using support VectorMachine

Author(s) : Danilo Dessi, Rim Helaoui, Vivek Kumar, Diego Reforgiato Recupero, Daniele Riboni

Links : PDF - Abstract

Code :
Coursera

Keywords : learning - dataset - results - deep - word -

Leave a Reply

Your email address will not be published. Required fields are marked *