Andrew Trask and David Gilmore, 2014 Belmont computer science graduates, will present their research at the 32nd International Conference on Machine Learning (ICML), the leading international machine learning conference, in France. Their paper titled, “Modeling Order in Neural Word Embeddings at Scale,” describes the deep neural network built at their employer, Digital Reasoning and is co-authored by Digital Reasoning’s Chief Technology Officer Matthew Russell. Neural Networks are computer systems that are modeled after the human brain and can gather new data, process it and react to it. The paper details both the impressive scope of their neural network as well as the exponential improvement in quality.
The design for the network is based on ideas Trask developed studying at Belmont. The parallel neural network is 14 times larger than the previous world record (built at Google), and performs 40 percent better in a key language-recognition benchmark than any other program. Their paper will be published in Volume 37 of the Journal of Machine Learning Research.
For more information, click here.
