CNN's are best known for their ability to recognize patterns present in images, and so the task chosen for the network described in this post was that of image classification. One of the most common benchmarks for gauging how well a computer vision algorithm performs is to train it on the MNIST handwritten digit database: a collection of 70,000 handwritten digits and their corresponding labels. The goal is to train a CNN to be as accurate as possible when labeling handwritten digits (ranging from 0–9). After about five hours of training and two loops over the training set, the network presented here was able to achieve an accuracy of 98% on the test data, meaning it could correctly guess almost every handwritten digit shown to it. Let's go over the individual components that form the network and how they link together to form predictions from the input data.

Abstract: The well-documented global shortage of radiologists is most acutely manifested in countries where the rapid rise of a middle class has created a new capacity to produce imaging studies at a rate which far exceeds the time required to train experts capable of interpreting such studies. The production to interpretation gap is seen clearly in the case of the most common of imaging studies: the chest x-ray, where technicians are increasingly called upon to not only acquire the image, but also to interpret it. The dearth of expert radiologists leads to both delayed and inaccurate diagnostic insights. The present study utilizes a robust radiology database, machine-learning technologies, and robust clinical validation to produce expert-level automatic interpretation of routine chest x-rays. Using a convolutional neural network (CNN) we achieve a performance which is slightly higher than radiologists in the detection of four common chest X-ray (CXR) findings which include focal lung opacities, diffuse lung opacity, cardiomegaly, and abnormal hilar prominence.

Francois is not only the creator of the Keras deep learning library, but he's also a Google AI researcher. He will also be speaking at PyImageConf 2018 in August of this year. Please, join me in welcoming Francois to the PyImageSearch -- it is truthfully a privilege to have him here. I know you are very busy with your work at Google AI and on the Keras library -- I really appreciate you taking the time to do this interview. It's quite the honor to have you on the PyImageSearch blog! For people who don't know you, who are you and what do you do?

This is accomplished by creating a computational graph, which takes multidimensional matrices called "Tensors" and does computations on them. When creating the graph, you have the possibility to explicitly specify where the computations should be done, on the GPU or CPU. By default it will check if a GPU is available, and use that. The Graph is run in a Session, where you specify what operations to execute in the run-function. Data from outside may also be supplied to placeholders in the graph, so you can run it multiple times with different input.

Recurrent Neural Nets (RNN) detect features in sequential data (e.g. Examples of applications which can be made using RNN's are anomaly detection in time-series data, classification of ECG and EEG data, stock market prediction, speech recogniton, sentiment analysis, etc. This is done by unrolling the data into N different copies of itself (if the data consists of N time-steps) . In this way, the input data at the previous time steps can be used when the data at timestep is evaluated. If the data at the previous time steps is somehow correlated to the data at the current time step, these correlations are remembered and otherwise they are forgotten.

Google DeepMind has proved pretty effective at mastering various games over the years. It is officially the finest player of Go in the world after beating the sitting human world champion and has since turned its figurative hand to Starcraft 2, while also proving adept at guessing when patients are going to die. It truly is a jack of all trades. The problem is that DeepMind is very much a lone wolf. It works alone and doesn't necessarily know how to play nice with humans – or indeed other AIs. You might not consider the best remedy for this to be arming DeepMind, but the good news is that the weapons and battlefield are purely virtual: the AI has learned how to play Quake III's Capture the Flag multiplayer mode and, predictably, has got to the point where it can teach humans a thing or two.

University of Wisconsin-Madison researchers found that a deep learning algorithm yielded similar sensitivity and specificity in differentiating benign and malignant bone lesions on F-18 NaF-PET/CT scans of patients with metastatic prostate cancer, but enabled significantly faster interpretation times compared with a traditional radiomics-based machine learning algorithm. The findings, presented at the SNMMI Annual Meeting, suggest that the deep learning algorithm may be "a great alternative to a radiomics model," said researcher Tyler Bradshaw.

A novel machine learning technique that would predict the biological age of a muscle and help combat sarcopenia has been developed. Sarcopenia is one of the major age-related processes and involves the loss of skeletal muscles and its function. The deep-learning based model can be used to estimate the relevant importance of the genetic and epigenetic factors driving this process within many age groups. "We believe that the most effective anti-ageing therapy should be tissue-specific, so we focused on the development of tissue-specific biomarkers of ageing. This work is an example of a marker of skeletal muscle tissue," said Polina Mamoshina, senior scientist at Insilico Medicine – a US-based next-generation artificial intelligence company.

Data-driven methods for improving turbulence modeling in Reynolds-Averaged Navier-Stokes (RANS) simulations have gained significant interest in the computational fluid dynamics community. Modern machine learning models have opened up a new area of black-box turbulence models allowing for the tuning of RANS simulations to increase their predictive accuracy. While several data-driven turbulence models have been reported, the quantification of the uncertainties introduced has mostly been neglected. Uncertainty quantification for such data-driven models is essential since their predictive capability rapidly declines as they are tested for flow physics that deviate from that in the training data. In this work, we propose a novel data-driven framework that not only improves RANS predictions but also provides probabilistic bounds for fluid quantities such as velocity and pressure. The uncertainties capture include both model form uncertainty as well as epistemic uncertainty induced by the limited training data. An invariant Bayesian deep neural network is used to predict the anisotropic tensor component of the Reynolds stress. This model is trained using Stein's variational gradient decent algorithm. The computed uncertainty on the Reynolds stress is propagated to the quantities of interest by vanilla Monte Carlo simulation. Results are presented for two test cases that differ geometrically from the training flows at several different Reynolds numbers. The prediction enhancement of the data-driven model is discussed as well as the associated probabilistic bounds for flow properties of interest. Ultimately this framework allows for a quantitative measurement of model confidence and uncertainty quantification for flows in which no high-fidelity observations or prior knowledge is available.

Neural networks use their hidden layers to transform input data into linearly separable data clusters, with a linear or a perceptron type output layer making the final projection on the line perpendicular to the discriminating hyperplane. For complex data with multimodal distributions this transformation is difficult to learn. Projection on $k\geq 2$ line segments is the simplest extension of linear separability, defining much easier goal for the learning process. Simple problems are 2-separable, but problems with inherent complex logic may be solved in a simple way by $k$-separable projections. The difficulty of learning non-linear data distributions is shifted to separation of line intervals, simplifying the transformation of data by hidden network layers. For classification of difficult Boolean problems, such as the parity problem, linear projection combined with \ksep is sufficient and provides a powerful new target for learning. More complex targets may also be defined, changing the goal of learning from linear discrimination to creation of data distributions that can easily be handled by specialized models selected to analyze output distributions. This approach can replace many layers of transformation required by deep learning models.