BELLEVUE, Wash., April 5, 2017 – DimensionalMechanics, a pioneer in deep learning engineering, today announced that the company has been selected to present a research paper at the 5th International Conference on Learning Representations (ICLR). The ICLR is a major industry event for artificial intelligence (AI) leaders taking place from April 24-26, 2017 at Palais des Congrès Neptune in Toulon, France.
DimensionalMechanics will present its paper, Charged Point Normalization: An Efficient Solution to the Saddle Point Problem, during a workshop track at the upcoming event. The paper introduces a dynamic type of normalization that forces the system to escape saddle points. Unlike other saddle point escaping algorithms, second order information is not utilized, and the system can be trained with an arbitrary gradient descent learner. The system drastically improves learning in a range of deep neural networks on various data-sets in comparison to non-CPN neural networks.
DimensionalMechanics enables organizations to build, control and deploy AI solutions, without in-house machine learning (ML) expertise. The company provides existing ML specialists the ability to work faster than ever before. DimensionalMechanics’ NeoPulse® Framework is an intuitive development and management platform for AI, offering a flexible suite of horizontal, reusable tools to meet the varying needs of organizations and users.
ICLR is an annual conference sponsored by the Computational and Biological Learning Society. 2017 marks the 5th annual ICLR, which is chaired by Yoshua Bengio, Head of the Montreal Institute for Learning Algorithms at Université de Montréal and Yann Lecun, Direcor of AI Research at Facebook and Professor at New York University. This year the show will focus on the rapidly developing field of representation learning, dealing with questions surrounding how to learn from meaningful and useful representations of data. The conference will include keynote presentations from invited speakers, oral presentations, and posters on topics such as deep learning and feature learning, metric learning, compositional modeling, structured prediction, reinforcement learning, and issues regarding large-scale learning and non-convex optimization.
DimensionalMechanics Inc®. was founded to make artificial intelligence (AI) accessible to organizations of all sizes. The company’s NeoPulse Framework is an intuitive development and management platform for AI, and includes a suite of horizontal, reusable tools that enable companies to answer vital questions using a simplified and consistent approach. NeoPulse also enables organization to distribute and monetize their AI creations. Based in Bellevue, WA, the company’s team brings decades of experience in the fields of machine learning and artificial intelligence from leading companies and research organizations.