Andrej Karpathy's post "The Unreasonable Effectiveness of Recurrent Neural Networks" made splashes last year. The basic premise is that you can create a recurrent neural network to learn language features character-by-character. But is the resultant model any different from a Markov chain built for the same purpose? I implemented a character-by-character Markov chain in R to find out. First, let's play a variation of the Imitation Game with generated text from Karpathy's tinyshakespeare dataset.

Kimberly Powell, who leads Nvidia's efforts in health care, says the company is working with medical researchers in a range of areas and will look to expand these efforts in coming years. Most notably, a machine-learning technique called deep learning is being applied to processing medical images and sifting through large amounts of medical data. Nvidia is, for example, working with Bradley Erickson, a neuro-radiologist at the Mayo Clinic, to apply deep learning to brain images. There are, however, significant challenges in applying techniques like deep learning to medicine.

A California-based startup called Instrumental developed an intelligent AI inspection system to help manufactures identify product defects on the assembly line. The California-based startup, founded by two form Apple engineers have raised over $10 million to make it easier to manufacture electronics and head off complicated problems before they start costing companies thousands of dollars a minute. Their customers, including Fortune 500 companies, have used the system to virtually disassemble 16,000 units and to take over 40,000 measurements, all remotely. Instrumental makes a hardware box that goes on the assembly line and takes a photo of every device that passes through and they recently announced their deep learning software called Detect which highlights units that appear defective or anomalous, giving our customers a significant edge in discovering and resolving product issues. Using TITAN X GPUs and cuDNN with the TensorFlow deep learning framework, they are able to process hundreds of units in seconds and identify the most interesting units to review.

This paper considers the use of Machine Learning (ML) in medicine by focusing on the main problem that this computational approach has been aimed at solving or at least minimizing: uncertainty. To this aim, we point out how uncertainty is so ingrained in medicine that it biases also the representation of clinical phenomena, that is the very input of ML models, thus undermining the clinical significance of their output. Recognizing this can motivate both medical doctors, in taking more responsibility in the development and use of these decision aids, and the researchers, in pursuing different ways to assess the value of these systems. In so doing, both designers and users could take this intrinsic characteristic of medicine more seriously and consider alternative approaches that do not "sweep uncertainty under the rug" within an objectivist fiction, which everyone can come up by believing as true.

This paper introduces the YouTube-8M Video Understanding Challenge hosted as a Kaggle competition and also describes my approach to experimenting with various models. For each of my experiments, I provide the score result as well as possible improvements to be made. Towards the end of the paper, I discuss the various ensemble learning techniques that I applied on the dataset which significantly boosted my overall competition score. At last, I discuss the exciting future of video understanding research and also the many applications that such research could significantly improve.

We are increasingly surrounded by artificially intelligent technology that takes decisions and executes actions on our behalf. This creates a pressing need for general means to communicate with, instruct and guide artificial agents, with human language the most compelling means for such communication. To achieve this in a scalable fashion, agents must be able to relate language to the world and to actions; that is, their understanding of language must be grounded and embodied. However, learning grounded language is a notoriously challenging problem in artificial intelligence research. Here we present an agent that learns to interpret language in a simulated 3D environment where it is rewarded for the successful execution of written instructions. Trained via a combination of reinforcement and unsupervised learning, and beginning with minimal prior knowledge, the agent learns to relate linguistic symbols to emergent perceptual representations of its physical surroundings and to pertinent sequences of actions. The agent's comprehension of language extends beyond its prior experience, enabling it to apply familiar language to unfamiliar situations and to interpret entirely novel instructions. Moreover, the speed with which this agent learns new words increases as its semantic knowledge grows. This facility for generalising and bootstrapping semantic knowledge indicates the potential of the present approach for reconciling ambiguous natural language with the complexity of the physical world.

We examine which top tools are "friends", their Python vs R bias, and which work well with Spark/Hadoop and Deep Learning, and identify an emerging Big Data Deep Learning ecosystem.

Artificial Intelligence is, locally, a computer algorithm tasked with solving input problems based on accessible data and operational parameters, with respect to the amount of computational power available to the algorithm. More generally, AI is the name given to machine intelligence. With the vast field of AI are specific concepts like machine learning and deep learning. In the same way as Russian Matryoshka dolls where the small doll is nested inside the bigger one, each of the three segments (Deep Learning, ML and AI) is a subset of the other. Advances in these three technologies are already revolutionizing many aspects of modern life, and although very much related, they are not the same.

As members of the deep learning R&D team at SVDS, we are interested in comparing Recurrent Neural Network (RNN) and other approaches to speech recognition. Until a few years ago, the state-of-the-art for speech recognition was a phonetic-based approach including separate components for pronunciation, acoustic, and language models. Typically, this consists of n-gram language models combined with Hidden Markov models (HMM). We wanted to start with this as a baseline model, and then explore ways to combine it with newer approaches such as Baidu's Deep Speech. While summaries exist explaining these baseline phonetic models, there do not appear to be any easily-digestible blog posts or papers that compare the tradeoffs of the different freely available tools.

Douglas Hofstadter in his book "I am a Strange Loop" coined this idea: Where he describes this self-referential mechanism as what describes the unique property of minds. The strange loop is a cyclic system that traverses several layers in a hierarchy. By moving through this cycle one finds oneself where one originally started. Coincidentally enough, this'strange loop' is in fact is the fundamental reason for what Yann LeCun describes as "the coolest idea in machine learning in the last twenty years." Loops are not typical in Deep Learning systems. These systems have conventionally been composed of acyclic graphs of computation layers.