End-to-end Fast Fourier Descriptors for Signal Authentication with Non-Coherent Points – We present a novel method for unsupervised neural architectures to encode and decode text into multiple time-space representations. The technique is based on a reinforcement learning algorithm and is evaluated on various real-world data sets. We demonstrate theoretically that the obtained temporal representations encode a rich syntactic dependency structure among the data, and encode a discriminative representation that learns to separate the syntactic dependency structure from the temporal structure. Our algorithm outperforms the state of the art baselines in both synthetic and real time.

We propose a new novel approach to the problem of learning natural dialogues via deep encoder-decoder neural networks (DCDNNs). First, we augment the deep convolutional network with two layers of DNNs: one for dialogue, and one for language. Next, we train the convolutional DCDNNs to learn a convolutional dictionary with various convolutional sub-problems. Our approach leverages both hand-crafted and annotated dialogues. We propose two models that outperform DCDNN-based models (both train the DCNNs to learn the dictionary). Finally, we experiment our approach on a large collection of 10,000 human and 100,000 character short dialogues. To evaluate our approach, we conduct a trial on an audience sample for the SemEval 2017 evaluation of a class of short dialogues with 2.2 million dialogues.

RoboJam: A Large Scale Framework for Multi-Label Image Monolingual Naming

Unsupervised Deep Learning With Shared Memory

End-to-end Fast Fourier Descriptors for Signal Authentication with Non-Coherent Points

A Neural Approach to Reinforcement Learning and Control of Scheduling Problems

Probabilistic Learning and Sparse Visual Saliency in Handwritten CharactersWe propose a new novel approach to the problem of learning natural dialogues via deep encoder-decoder neural networks (DCDNNs). First, we augment the deep convolutional network with two layers of DNNs: one for dialogue, and one for language. Next, we train the convolutional DCDNNs to learn a convolutional dictionary with various convolutional sub-problems. Our approach leverages both hand-crafted and annotated dialogues. We propose two models that outperform DCDNN-based models (both train the DCNNs to learn the dictionary). Finally, we experiment our approach on a large collection of 10,000 human and 100,000 character short dialogues. To evaluate our approach, we conduct a trial on an audience sample for the SemEval 2017 evaluation of a class of short dialogues with 2.2 million dialogues.