Hospital Readmissions within 30 days after discharge following Coronary Artery Bypass Graft (CABG) Surgery are substantial contributors to healthcare costs. Many predictive models were developed to identify risk factors for readmissions. However, majority of the existing models use statistical analysis techniques with data available at discharge. We propose an ensembled model to predict CABG readmissions using pre-discharge perioperative data and machine learning survival analysis techniques. Firstly, we applied fifty one potential readmission risk variables to Cox Proportional Hazard (CPH) survival regression univariate analysis. Fourteen of them turned out to be significant (with p value < 0.05), contributing to readmissions. Subsequently, we applied these 14 predictors to multivariate CPH model and Deep Learning Neural Network (NN) representation of the CPH model, DeepSurv. We validated this new ensembled model with 453 isolated adult CABG cases. Nine of the fourteen perioperative risk variables were identified as the most significant with Hazard Ratios (HR) of greater than 1.0. The concordance index metrics for CPH, DeepSurv, and ensembled models were then evaluated with training and validation datasets. Our ensembled model yielded promising results in terms of c-statistics, as we raised the the number of iterations and data set sizes. 30 day all-cause readmissions among isolated CABG patients can be predicted more effectively with perioperative pre-discharge data, using machine learning survival analysis techniques. Prediction accuracy levels could be improved further with deep learning algorithms.

Neural networks currently dominate the machine learning community and they do so for good reasons. Their accuracy on complex tasks such as image classification is unrivaled at the moment and with recent improvements they are reasonably easy to train. Nevertheless, neural networks are lacking robustness and interpretability. Prototype-based vector quantization methods on the other hand are known for being robust and interpretable. For this reason, we propose techniques and strategies to merge both approaches. This contribution will particularly highlight the similarities between them and outline how to construct a prototype-based classification layer for multilayer networks. Additionally, we provide an alternative, prototype-based, approach to the classical convolution operation. Numerical results are not part of this report, instead the focus lays on establishing a strong theoretical framework. By publishing our framework and the respective theoretical considerations and justifications before finalizing our numerical experiments we hope to jump-start the incorporation of prototype-based learning in neural networks and vice versa.

Generative models that learn to associate variations in the output along isolated attributes with coordinate directions in latent space are said to possess disentangled representations. This has been a recent topic of great interest, but remains poorly understood. We show that even for GANs that do not possess disentangled representations, one can find paths in latent space over which local disentanglement occurs. These paths are determined by the leading right-singular vectors of the Jacobian of the generator with respect to its input. We describe an efficient regularizer that aligns these vectors with the coordinate axes, and show that it can be used to induce high-quality, disentangled representations in GANs, in a completely unsupervised manner.

Abstract-- Sampling-based motion planning is an effective tool to compute safe trajectories for automated vehicles in complex environments. However, a fast convergence to the optimal solution can only be ensured with the use of problemspecific samplingdistributions. Due to the large variety of driving situations within the context of automated driving, it is very challenging to manually design such distributions. This paper introduces therefore a data-driven approach utilizing a deep convolutional neural network (CNN): Given the current driving situation, future ego-vehicle poses can be directly generated from the output of the CNN allowing to guide the motion planner efficiently towards the optimal solution. A benchmark highlights that the CNN predicts future vehicle poses with a higher accuracy compared to uniform sampling and a state-of-the-art A*-based approach. Combining this CNNguided samplingwith the motion planner Bidirectional RRT* reduces the computation time by up to an order of magnitude and yields a faster convergence to a lower cost as well as a success rate of 100 % in the tested scenarios. I. INTRODUCTION Motion planning is one of the major pillars in the software architecture of automated vehicles. Its task is to compute a safe trajectory from start goal taking into account the vehicle's constraints, the non-convex surrounding as well as the comfort requirements of passengers. In structured environments, such as highway driving, it is sufficient to solve the motion planning problem locally close to the lane centerline.

We view molecular optimization as a graph-to-graph translation problem. The goal is to learn to map from one molecular graph to another with better properties based on an available corpus of paired molecules. Since molecules can be optimized in different ways, there are multiple viable translations for each input graph. A key challenge is therefore to model diverse translation outputs. Our primary contributions include a junction tree encoder-decoder for learning diverse graph translations along with a novel adversarial training method for aligning distributions of molecules. Diverse output distributions in our model are explicitly realized by low-dimensional latent vectors that modulate the translation process. We evaluate our model on multiple molecular optimization tasks and show that our model outperforms previous state-of-the-art baselines by a significant margin.

In complex transfer learning scenarios new tasks might not be tightly linked to previous tasks. Approaches that transfer information contained only in the final parameters of a source model will therefore struggle. Instead, transfer learning at a higher level of abstraction is needed. We propose Leap, a framework that achieves this by transferring knowledge across learning processes. We associate each task with a manifold on which the training process travels from initialization to final parameters and construct a meta learning objective that minimizes the expected length of this path. Our framework leverages only information obtained during training and can be computed on the fly at negligible cost. We demonstrate that our framework outperforms competing methods, both in meta learning and transfer learning, on a set of computer vision tasks. Finally, we demonstrate that Leap can transfer knowledge across learning processes in demanding Reinforcement learning environments (Atari) that involve millions of gradient steps.

Learning is an inherently continuous phenomenon. When humans learn a new task there is no explicit distinction between training and inference. After we learn a task, we keep learning about it while performing the task. What we learn and how we learn it varies during different stages of learning. Learning how to learn and adapt is a key property that enables us to generalize effortlessly to new settings. This is in contrast with conventional settings in machine learning where a trained model is frozen during inference. In this paper we study the problem of learning to learn at both training and inference time in the context of visual navigation. A fundamental challenge in navigation is generalization to unseen scenes. In this paper we propose a self-adaptive visual navigation method (SAVN) which learns to adapt to new environments without any explicit supervision. Our solution is a meta-reinforcement learning approach where an agent learns a self-supervised interaction loss that encourages effective navigation. Our experiments, performed in the AI2-THOR framework, show major improvements in both success rate and SPL for visual navigation in novel scenes.

De novo protein structure prediction from amino acid sequence is one of the most challenging problems in computational biology. As one of the extensively explored mathematical models for protein folding, Hydrophobic-Polar (HP) model enables thorough investigation of protein structure formation and evolution. Although HP model discretizes the conformational space and simplifies the folding energy function, it has been proven to be an NP-complete problem. In this paper, we propose a novel protein folding framework FoldingZero, self-folding a de novo protein 2D HP structure from scratch based on deep reinforcement learning. FoldingZero features the coupled approach of a two-head (policy and value heads) deep convolutional neural network (HPNet) and a regularized Upper Confidence Bounds for Trees (R-UCT). It is trained solely by a reinforcement learning algorithm, which improves HPNet and R-UCT iteratively through iterative policy optimization. Without any supervision and domain knowledge, FoldingZero not only achieves comparable results, but also learns the latent folding knowledge to stabilize the structure. Without exponential computation, FoldingZero shows promising potential to be adopted for real-world protein properties prediction.

Knowledge distillation is an effective technique that transfers knowledge from a large teacher model to a shallow student. However, just like massive classification, large scale knowledge distillation also imposes heavy computational costs on training models of deep neural networks, as the softmax activations at the last layer involve computing probabilities over numerous classes. In this work, we apply the idea of importance sampling which is often used in Neural Machine Translation on large scale knowledge distillation. W e present a method called dynamic importance sampling, where ranked classes are sampled from a dynamic distribution derived from the interaction between the teacher and student in full distillation. W e highlight the utility of our proposal prior which helps the student capture the main information in the loss function. Our approach manages to reduce the computational cost at training time while maintaining the competitive performance on CIF AR-100 and Market-1501 person re-identification datasets.

The model reduction problem that eases the computation costs and latency of complex deep learning architectures has received an increasing number of investigations owing to its importance in model deployment. One promising method is knowledge distillation (KD), which creates a fast-to-execute student model to mimic a large teacher network. In this paper, we propose a method, called KDFM (Knowledge Distillation with Feature Maps), which improves the effectiveness of KD by learning the feature maps from the teacher network. Two major techniques used in KDFM are shared classifier and generative adversarial network. Experimental results show that KDFM can use a four layers CNN to mimic DenseNet-40 and use MobileNet to mimic DenseNet-100. Both student networks have less than 1\% accuracy loss comparing to their teacher models for CIFAR-100 datasets. The student networks are 2-6 times faster than their teacher models for inference, and the model size of MobileNet is less than half of DenseNet-100's.