In recent years, deep neural networks have been applied to obtain high performance of prediction, classification, and pattern recognition. However, the weights in these deep neural networks are difficult to be explained. Although a linear regression method can provide explainable results, the method is not suitable in the case of input interaction. Therefore, an explainable deep neural network (ExpDNN) with explainable layers is proposed to obtain explainable results in the case of input interaction. Three cases were given to evaluate the proposed ExpDNN, and the results showed that the absolute value of weight in an explainable layer can be used to explain the weight of corresponding input for feature extraction.

Inspired by recent successes with parallel optimization techniques for solving Boolean satisfiability, we investigate a set of strategies and heuristics that aim to leverage parallel computing to improve the scalability of neural network verification. We introduce an algorithm based on partitioning the verification problem in an iterative manner and explore two partitioning strategies, that work by partitioning the input space or by case splitting on the phases of the neuron activations, respectively. We also introduce a highly parallelizable pre-processing algorithm that uses the neuron activation phases to simplify the neural network verification problems. An extensive experimental evaluation shows the benefit of these techniques on both existing benchmarks and new benchmarks from the aviation domain. A preliminary experiment with ultra-scaling our algorithm using a large distributed cloud-based platform also shows promising results.

A new AI system developed by researchers from Google and the University of California could prevent deaths caused by incorrect prescriptions. While quite rare, prescriptions that are incorrect – or react badly to a patient's existing medications – can result in hospitalisation or even death. In a blog post today, Alvin Rajkomar MD, Research Scientist and Eyal Oren PhD, Product Manager, Google AI, set out their work on using AI for medical predictions. The AI is able to predict which conditions a patient is being treated for based on certain parameters. "For example, if a doctor prescribed ceftriaxone and doxycycline for a patient with an elevated temperature, fever and cough, the model could identify these as signals that the patient was being treated for pneumonia," the researchers wrote.

Ever caught yourself wondering what artificial intelligence (AI), machine learning(ML), and deep learning (DL) are and how they are different? Well, let me tell you that you are not alone. These three terms are so frequently used together, dare I say, rather interchangeably that one begins to think they are all the same. No, they are simply not. However, they are very closely related.

Artificial intelligence (AI) is quickly becoming a critical component in how government, business and citizens defend themselves against cyber attacks. Starting with technology designed to automate specific manual tasks, and advancing to machine learning using increasingly complex systems to parse data, breakthroughs in deep learning capabilities will become an integral part of the security agenda. Much attention is paid to how these capabilities are helping to build a defence posture. But how enemies might harness AI to drive a new generation of attack vectors, and how the community might respond, is often overlooked. Ultimately, the real danger of AI lies in how it will enable attackers.

The neural network method of solving differential equations is used to approximate the electric potential and corresponding electric field in the slit-well microfluidic device. The device's geometry is non-convex, making this a challenging problem to solve using the neural network method. To validate the method, the neural network solutions are compared to a reference solution obtained using the finite element method. Additional metrics are presented that measure how well the neural networks recover important physical invariants that are not explicitly enforced during training: spatial symmetries and conservation of electric flux. Finally, as an application-specific test of validity, neural network electric fields are incorporated into particle simulations. Conveniently, the same loss functional used to train the neural networks also seems to provide a reliable estimator of the networks' true errors, as measured by any of the metrics considered here. In all metrics, deep neural networks significantly outperform shallow neural networks, even when normalized by computational cost. Altogether, the results suggest that the neural network method can reliably produce solutions of acceptable accuracy for use in subsequent physical computations, such as particle simulations.

Subsurface flow problems usually involve some degree of uncertainty. Consequently, uncertainty quantification is commonly necessary for subsurface flow prediction. In this work, we propose a methodology for efficient uncertainty quantification for dynamic subsurface flow with a surrogate constructed by the Theory-guided Neural Network (TgNN). The TgNN here is specially designed for problems with stochastic parameters. In the TgNN, stochastic parameters, time and location comprise the input of the neural network, while the quantity of interest is the output. The neural network is trained with available simulation data, while being simultaneously guided by theory (e.g., the governing equation, boundary conditions, initial conditions, etc.) of the underlying problem. The trained neural network can predict solutions of subsurface flow problems with new stochastic parameters. With the TgNN surrogate, the Monte Carlo (MC) method can be efficiently implemented for uncertainty quantification. The proposed methodology is evaluated with two-dimensional dynamic saturated flow problems in porous medium. Numerical results show that the TgNN based surrogate can significantly improve the efficiency of uncertainty quantification tasks compared with simulation based implementation. Further investigations regarding stochastic fields with smaller correlation length, larger variance, changing boundary values and out-of-distribution variances are performed, and satisfactory results are obtained.

The ability to recognize objects despite there being differences in appearance, known as Core Object Recognition, forms a critical part of human perception. While it is understood that the brain accomplishes Core Object Recognition through feedforward, hierarchical computations through the visual stream, the underlying algorithms that allow for invariant representations to form downstream is still not well understood. (DiCarlo et al., 2012) Various computational perceptual models have been built to attempt and tackle the object identification task in an artificial perceptual setting. Artificial Neural Networks, computational graphs consisting of weighted edges and mathematical operations at vertices, are loosely inspired by neural networks in the brain and have proven effective at various visual perceptual tasks, including object characterization and identification. (Pinto et al., 2008) (DiCarlo et al., 2012) For many data analysis tasks, learning representations where each dimension is statistically independent and thus disentangled from the others is useful. If the underlying generative factors of the data are also statistically independent, Bayesian inference of latent variables can form disentangled representations. This thesis constitutes a research project exploring a generalization of the Variational Autoencoder (VAE), $\beta$-VAE, that aims to learn disentangled representations using variational inference. $\beta$-VAE incorporates the hyperparameter $\beta$, and enforces conditional independence of its bottleneck neurons, which is in general not compatible with the statistical independence of latent variables. This text examines this architecture, and provides analytical and numerical arguments, with the goal of demonstrating that this incompatibility leads to a non-monotonic inference performance in $\beta$-VAE with a finite optimal $\beta$.

In this work, orientation detection using Deep Learning is acknowledged for a particularly vulnerable class of road users,the cyclists. Knowing the cyclists' orientation is of great relevance since it provides a good notion about their future trajectory, which is crucial to avoid accidents in the context of intelligent transportation systems. Using Transfer Learning with pre-trained models and TensorFlow, we present a performance comparison between the main algorithms reported in the literature for object detection,such as SSD, Faster R-CNN and R-FCN along with MobilenetV2, InceptionV2, ResNet50, ResNet101 feature extractors. Moreover, we propose multi-class detection with eight different classes according to orientations. To do so, we introduce a new dataset called "Detect-Bike", containing 20,229 cyclist instances over 11,103 images, which has been labeled based on cyclist's orientation. Then, the same Deep Learning methods used for detection are trained to determine the target's heading. Our experimental results and vast evaluation showed satisfactory performance of all of the studied methods for the cyclists and their orientation detection, especially using Faster R-CNN with ResNet50 proved to be precise but significantly slower. Meanwhile, SSD using InceptionV2 provided good trade-off between precision and execution time, and is to be preferred for real-time embedded applications.

Comparing competing mathematical models of complex natural processes is a shared goal among many branches of science. The Bayesian probabilistic framework offers a principled way to perform model comparison and extract useful metrics for guiding decisions. However, many interesting models are intractable with standard Bayesian methods, as they lack a closed-form likelihood function or the likelihood is computationally too expensive to evaluate. With this work, we propose a novel method for performing Bayesian model comparison using specialized deep learning architectures. Our method is purely simulation-based and circumvents the step of explicitly fitting all alternative models under consideration to each observed dataset. Moreover, it involves no hand-crafted summary statistics of the data and is designed to amortize the cost of simulation over multiple models and observable datasets. This makes the method applicable in scenarios where model fit needs to be assessed for a large number of datasets, so that per-dataset inference is practically infeasible. Finally, we propose a novel way to measure epistemic uncertainty in model comparison problems. We argue that this measure of epistemic uncertainty provides a unique proxy to quantify absolute evidence even in a framework which assumes that the true data-generating model is within a finite set of candidate models.