Abstract: It is widely known that neural networks (NNs) are universal approximators of continuous functions, however, a less known but powerful result is that a NN with a single hidden layer can approximate accurately any nonlinear continuous operator. This universal approximation theorem of operators is suggestive of the potential of NNs in learning from scattered data any continuous operator or complex system. To realize this theorem, we design a new NN with small generalization error, the deep operator network (DeepONet), consisting of a NN for encoding the discrete input function space (branch net) and another NN for encoding the domain of the output functions (trunk net). We demonstrate that DeepONet can learn various explicit operators, e.g., integrals and fractional Laplacians, as well as implicit operators that represent deterministic and stochastic differential equations. We study, in particular, different formulations of the input function space and its effect on the generalization error.

This seminar talk will be hosted remotely via Zoom.

Zoom Webinar link:https://mit.zoom.us/j/95815924103?pwd=Y0Zrd3hiQWdGN3k3SlVORFJFZkRwUT09 

passcode 829729

Please note that this talk has been canceled.

We will reschedule his talk at the earliest convenience.

Abstract: Hypernetworks, also known as dynamic networks, are neural networks in which the weights of at least some of the layers vary dynamically based on the input. Such networks have composite architectures in which one network predicts the weights of another network. I will briefly describe the early days of dynamic layers and present recent results from diverse domains: 3D reconstruction from a single image, image retouching, electrical circuit design, decoding block codes, graph hypernetworks for bioinformatics, and action recognition in video. Finally, I will present a new hypernetwork-based model for the role of feedback in neural computations.

Title: Feedforward and feedback processes in visual recognition

Abstract: Progress in deep learning has spawned great successes in many engineering applications. As a prime example, convolutional neural networks, a type of feedforward neural networks, are now approaching – and sometimes even surpassing – human accuracy on a variety of visual recognition tasks. In this talk, however, I will show that these neural networks and their recent extensions exhibit a limited ability to solve seemingly simple visual reasoning problems involving incremental grouping, similarity and spatial relation judgments. Our group has developed a recurrent network model of classical and extra-classical receptive fields that is constrained by the anatomy and physiology of the visual cortex. The model was shown to account for diverse visual illusions providing computational evidence for a novel canonical circuit that is shared across visual modalities. I will show that this computational neuroscience model can be turned into a modern end-to-end trainable deep recurrent network architecture which addresses some of the shortcomings exhibited by state-of-the-art feedforward networks for solving complex visual reasoning tasks. This suggests that neuroscience may contribute powerful new ideas and approaches to computer science and artificial intelligence.​

Abstract: Jack Hidary will take us through the nascent, but promising field of quantum computing and his new book, Quantum Computing: An Applied Approach

Bio: Jack D. Hidary is a research scientist in quantum computing and in AI at Alphabet X, formerly Google X. He and his group develop and research algorithms for NISQ-regime quantum processors as well as create new software libraries for quantum computing.  In the  AI field, Jack and his  group focus on fundamental research such as the generalization of deep networks as well as applied AI technologies.