It aims to provide students with an understanding of the role computation can play in solving problems and to help students, regardless of their major, feel justifiably confident of their ability to write small programs that allow them to accomplish useful goals. The class uses the Python 3.5 programming language.

Parallel tempering addresses some of the drawbacks of canonical Markov Chain Monte-Carlo methods for Bayesian neural learning with the ability to utilize high performance computing. However, certain challenges remain given the large range of network parameters and big data. Surrogate-assisted optimization considers the estimation of an objective function for models given computational inefficiency or difficulty to obtain clear results. We address the inefficiency of parallel tempering for large-scale problems by combining parallel computing features with surrogate assisted estimation of likelihood function that describes the plausibility of a model parameter value, given specific observed data. In this paper, we present surrogate-assisted parallel tempering for Bayesian neural learning where the surrogates are used to estimate the likelihood. The estimation via the surrogate becomes useful rather than evaluating computationally expensive models that feature large number of parameters and datasets. Our results demonstrate that the methodology significantly lowers the computational cost while maintaining quality in decision making using Bayesian neural learning. The method has applications for a Bayesian inversion and uncertainty quantification for a broad range of numerical models.

Abstract--Crude oil is a major component in most advanced economies of the world. Accurately predicting and understanding thebehavior of crude oil prices is important for economists, analysts, forecasters, and traders, to name a few. The price of crude oil has declined in the past decade and is seeing a phase of stability; but will this stability last? This work is an empirical study on how multivariate analysis may be employed to predict crude oil spot prices using neural networks. The concept of using neural networks showed promising potential. A very simple neural network model was able to perform on par with ARIMA models - the state-of-the-art model in time-series forecasting. Advanced neural network models using larger datasets may be used in the future to extend this proofof-concept toa full scale framework. I. INTRODUCTION Crude oil spot prices saw a tremendous uptick in the first decade of the 21 Since 2014, crude oil prices have fallen and may have stabilized now. However, there has always been a constant interest in accurately predicting crude oil prices; given that crude oil drives a major portion of the economy. Economists, scientists, data analysts, and traders are all interested in models that give them the best accuracy.

One of the most important applications of seismic reflection is the hydrocarbon exploration which is closely related to salt deposits analysis. This problem is very important even nowadays due to it's non-linear nature. Taking into account the recent developments in deep learning networks TGS-NOPEC Geophysical Company hosted the Kaggle competition for salt deposits segmentation problem in seismic image data. In this paper, we demonstrate the great performance of several novel deep learning techniques merged into a single neural network which achieved the 27th place (top 1%) in the mentioned competition. Using a U-Net with ResNeXt-50 encoder pre-trained on ImageNet as our base architecture, we implemented Spatial-Channel Squeeze & Excitation, Lovasz loss, CoordConv and Hypercolumn methods. The source code for our solution is made publicly available at https://github.com/K-Mike/Automatic-salt-deposits-segmentation.

Over recent years, deep learning has become a mainstream method in machine learning. More advanced networks are being actively developed to solve real-world problems in many important areas. Among successful features of network architectures, shortcut connections are well established to take the outputs of earlier layers as the inputs to later layers, and produce excellent results such as in ResNet and DenseNet. Despite the power of shortcuts, there remain important questions on the underlying mechanism and associated functionalities. For example, will adding shortcuts lead to a more compact structure? How to use shortcuts for an optimal efficiency and capacity of the network model? Along this direction, here we demonstrate that given only one neuron in each layer, the shortcuts can be sparsely placed to let the slim network become an universal approximator. Potentially, our theoretically-guaranteed sparse network model can achieve a learning performance comparable to densely-connected networks on well-known benchmarks.

Transfer learning, which allows a source task to affect the inductive bias of the target task, is widely used in computer vision. The typical way of conducting transfer learning with deep neural networks is to fine-tune a model pre-trained on the source task using data from the target task. In this paper, we propose an adaptive fine-tuning approach, called SpotTune, which finds the optimal fine-tuning strategy per instance for the target data. In SpotTune, given an image from the target task, a policy network is used to make routing decisions on whether to pass the image through the fine-tuned layers or the pre-trained layers. We conduct extensive experiments to demonstrate the effectiveness of the proposed approach. Our method outperforms the traditional fine-tuning approach on 12 out of 14 standard datasets.We also compare SpotTune with other state-of-the-art fine-tuning strategies, showing superior performance. On the Visual Decathlon datasets, our method achieves the highest score across the board without bells and whistles.

Deep neural networks have shown remarkable performance across a wide range of vision-based tasks, particularly due to the availability of large-scale datasets for training and better architectures. However, data seen in the real world are often affected by distortions that not accounted for by the training datasets. In this paper, we address the challenge of robustness and stability of neural networks and propose a general training method that can be used to make the existing neural network architectures more robust and stable to input visual perturbations while using only available datasets for training. Proposed training method is convenient to use as it does not require data augmentation or changes in the network architecture. We provide theoretical proof as well as empirical evidence for the efficiency of the proposed training method by performing experiments with existing neural network architectures and demonstrate that same architecture when trained with the proposed training method perform better than when trained with conventional training approach in the presence of noisy datasets.

Deep neural networks are traditionally trained using human-designed stochastic optimization algorithms, such as SGD and Adam. Recently, the approach of learning to optimize network parameters has emerged as a promising research topic. However, these learned black-box optimizers sometimes do not fully utilize the experience in human-designed optimizers, therefore have limitation in generalization ability. In this paper, a new optimizer, dubbed as \textit{HyperAdam}, is proposed that combines the idea of "learning to optimize" and traditional Adam optimizer. Given a network for training, its parameter update in each iteration generated by HyperAdam is an adaptive combination of multiple updates generated by Adam with varying decay rates. The combination weights and decay rates in HyperAdam are adaptively learned depending on the task. HyperAdam is modeled as a recurrent neural network with AdamCell, WeightCell and StateCell. It is justified to be state-of-the-art for various network training, such as multilayer perceptron, CNN and LSTM.

Scalable Bayesian sampling is playing an important role in modern machine learning, especially in the fast-developed unsupervised-(deep)-learning models. While tremendous progresses have been achieved via scalable Bayesian sampling such as stochastic gradient MCMC (SG-MCMC) and Stein variational gradient descent (SVGD), the generated samples are typically highly correlated. Moreover, their sample-generation processes are often criticized to be inefficient. In this paper, we propose a novel self-adversarial learning framework that automatically learns a conditional generator to mimic the behavior of a Markov kernel (transition kernel). High-quality samples can be efficiently generated by direct forward passes though a learned generator. Most importantly, the learning process adopts a self-learning paradigm, requiring no information on existing Markov kernels, e.g., knowledge of how to draw samples from them. Specifically, our framework learns to use current samples, either from the generator or pre-provided training data, to update the generator such that the generated samples progressively approach a target distribution, thus it is called self-learning. Experiments on both synthetic and real datasets verify advantages of our framework, outperforming related methods in terms of both sampling efficiency and sample quality.

Integrated gradients are widely employed to evaluate the contribution of input features in classification models because it satisfies the axioms for attribution of prediction. This method, however, requires an appropriate baseline for reliable determination of the contributions. We propose a compensated integrated gradients method that does not require a baseline. In fact, the method compensates the attributions calculated by integrated gradients at an arbitrary baseline using Shapley sampling. We prove that the method retrieves reliable attributions if the processes of input features in a classifier are mutually independent, and they are identical like shared weights in convolutional neural networks. Using three electroencephalogram datasets, we experimentally demonstrate that the attributions of the proposed method are more reliable than those of the original integrated gradients, and its computational complexity is much lower than that of Shapley sampling.