This work introduces a method to tune a sequence-based generative model for molecular de novo design that through augmented episodic likelihood can learn to generate structures with certain specified desirable properties. We demonstrate how this model can execute a range of tasks such as generating analogues to a query structure and generating compounds predicted to be active against a biological target. As a proof of principle, the model is first trained to generate molecules that do not contain sulphur. As a second example, the model is trained to generate analogues to the drug Celecoxib, a technique that could be used for scaffold hopping or library expansion starting from a single molecule. Finally, when tuning the model towards generating compounds predicted to be active against the dopamine receptor type 2, the model generates structures of which more than 95% are predicted to be active, including experimentally confirmed actives that have not been included in either the generative model nor the activity prediction model.

Deep Recurrent Neural Networks (RNNs) achieve state-of-the-art results in many sequence-to-sequence tasks. However, deep RNNs are difficult to train and suffer from overfitting. We introduce a training method that trains the network gradually, and treats each layer individually, to achieve improved results in language modelling tasks. Training deep LSTM with Gradual Learning (GL) obtains perplexity of 61.7 on the Penn Treebank (PTB) corpus. As far as we know (as for the 20.05.2017), GL improves the best state-of-the-art performance by a single LSTM/RHN model on the word-level PTB dataset.

The recently proposed Multi-Layer Convolutional Sparse Coding (ML-CSC) model, consisting of a cascade of convolutional sparse layers, provides a new interpretation of Convolutional Neural Networks (CNNs). Under this framework, the computation of the forward pass in a CNN is equivalent to a pursuit algorithm aiming to estimate the nested sparse representation vectors -- or feature maps -- from a given input signal. Despite having served as a pivotal connection between CNNs and sparse modeling, a deeper understanding of the ML-CSC is still lacking: there are no pursuit algorithms that can serve this model exactly, nor are there conditions to guarantee a non-empty model. While one can easily obtain signals that approximately satisfy the ML-CSC constraints, it remains unclear how to simply sample from the model and, more importantly, how one can train the convolutional filters from real data. In this work, we propose a sound pursuit algorithm for the ML-CSC model by adopting a projection approach. We provide new and improved bounds on the stability of the solution of such pursuit and we analyze different practical alternatives to implement this in practice. We show that the training of the filters is essential to allow for non-trivial signals in the model, and we derive an online algorithm to learn the dictionaries from real data, effectively resulting in cascaded sparse convolutional layers. Last, but not least, we demonstrate the applicability of the ML-CSC model for several applications in an unsupervised setting, providing competitive results. Our work represents a bridge between matrix factorization, sparse dictionary learning and sparse auto-encoders, and we analyze these connections in detail.

Large-scale deep neural networks (DNNs) are both compute and memory intensive. As the size of DNNs continues to grow, it is critical to improve the energy efficiency and performance while maintaining accuracy. For DNNs, the model size is an important factor affecting performance, scalability and energy efficiency. Weight pruning achieves good compression ratios but suffers from three drawbacks: 1) the irregular network structure after pruning; 2) the increased training complexity; and 3) the lack of rigorous guarantee of compression ratio and inference accuracy. To overcome these limitations, this paper proposes CirCNN, a principled approach to represent weights and process neural networks using block-circulant matrices. CirCNN utilizes the Fast Fourier Transform (FFT)-based fast multiplication, simultaneously reducing the computational complexity (both in inference and training) from O(n2) to O(nlogn) and the storage complexity from O(n2) to O(n), with negligible accuracy loss. Compared to other approaches, CirCNN is distinct due to its mathematical rigor: it can converge to the same effectiveness as DNNs without compression. The CirCNN architecture, a universal DNN inference engine that can be implemented on various hardware/software platforms with configurable network architecture. To demonstrate the performance and energy efficiency, we test CirCNN in FPGA, ASIC and embedded processors. Our results show that CirCNN architecture achieves very high energy efficiency and performance with a small hardware footprint. Based on the FPGA implementation and ASIC synthesis results, CirCNN achieves 6-102X energy efficiency improvements compared with the best state-of-the-art results.

Movidius and Intel have put deep-learning on a stick with a tiny $79 USB device that makes bringing AI to hardware a snap. In April of last year, Movidius showed off the first iteration of this device, which they then called the Fathom Neural Compute Stick. The company wasn't able to get the product out as quickly as they had hoped because they were a little busy getting acquired by Intel. Movidius's goal has long been to move this type of image-based deep learning from the cloud to the edge with its Myriad 2 visual processing unit. The chips are used on everything from security cameras and drones to AR headsets, enabling them all to recognize and identify objects in the world around them.

– Big data helps to make strategy for future and understand user behaviors. In 1959, Arther Samuel gave very simple definition of Machine Learning as "a Field of study that gives computer the ability to learn without being explicitly programmed". Now almost after 58 years from then we still have not progressed much beyond this definition if we compare the progress we made in other areas from same time. The idea of Fintech adopting some best practices from the Big Data and AI (Artificial Intelligence, Machine Learning and Deep Learning) is not so new, have you heard of accepting selfie as authentication for your shopping bill payment, Siri on your iPhone etc. A Decentralized Autonomous Organization (DAO) is a process that manifests these characteristics.

PUNE: California-headquartered data analytics firm Saama Technologies is setting up a deep learning centre in Pune as it attempts to build on its expertise in the data analytics space. Ken Coleman, chairman, Saama Technologies, told ET that the company was in the process of hiring people for the lab, which would work in collaboration with its existing engineering centre in the city. "Pune will play an increasingly important role in our success. We are not in India for the low costs but to tap into the brain power," said Coleman, who is also an advisor to venture capital firm Andreessen Horowitz. Saama would also be increasing its headcount by 15-20% by the end of this year, he said.

Nevertheless, we are no longer talking about a traditional, instantly available analytics packages. In order for you, as well, to experience the benefits of the IoT, you are required to efficiently manage the volume, velocity, and variety of the data, employ predictive analytics to model IoT device performance properly and to adopt a predictive strategy to monitor, optimize and maintain your IoT offerings. According to Picnet IT consulting company, for utmost results you need advanced machine learning and deep learning technologies, joined in a streamlined delivery framework which will provide accurate answers about the business conditions of the future, which will, as a result, improve your decision-making and risk-assessment skills.

Over the past few years, much of the progress in deep learning for computer vision can be boiled down to just a handful of neural network architectures. Setting aside all the math, the code, and the implementation details, I wanted to explore one simple question: how and why do these models work? The VGG networks, along with the earlier AlexNet from 2012, follow the now archetypal layout of basic conv nets: a series of convolutional, max-pooling, and activation layers before some fully-connected classification layers at the end. MobileNet is essentially a streamlined version of the Xception architecture optimized for mobile applications. The remaining three, however, truly redefine the way we look at neural networks.

Can we learn something from the AI playing against Arteezy? I also give a conclusion about what the bot is going to be doing in the future and how OpenAI will try to achieve this.