Acquiring ground truth labels for unlabelled data can be a costly procedure, since it often requires manual labour that is error-prone. Consequently, the available amount of labelled data is increasingly reduced due to the limitations of manual data labelling. It is possible to increase the amount of labelled data samples by performing automated labelling or crowd-sourcing the annotation procedure. However, they often introduce noise or uncertainty in the labelset, that leads to decreased performance of supervised deep learning methods. On the other hand, weak supervision methods remain robust during noisy labelsets or can be effective even with low amounts of labelled data. In this paper we evaluate the effectiveness of a representation learning method that uses external categorical evidence called "Evidence Transfer", against low amount of corresponding evidence termed as incomplete evidence. Evidence transfer is a robust solution against external unknown categorical evidence that can introduce noise or uncertainty. In our experimental evaluation, evidence transfer proves to be effective and robust against different levels of incompleteness, for two types of incomplete evidence.

Deep neural networks have gained tremendous success in a broad range of machine learning tasks due to its remarkable capability to learn semantic-rich features from high-dimensional data. However, they often require large-scale labelled data to successfully learn such features, which significantly hinders their adaption into unsupervised learning tasks, such as anomaly detection and clustering, and limits their applications into critical domains where obtaining massive labelled data is prohibitively expensive. To enable downstream unsupervised learning on those domains, in this work we propose to learn features without using any labelled data by training neural networks to predict data distances in a randomly projected space. Random mapping is a theoretical proven approach to obtain approximately preserved distances. To well predict these random distances, the representation learner is optimised to learn genuine class structures that are implicitly embedded in the randomly projected space. Experimental results on 19 real-world datasets show our learned representations substantially outperform state-of-the-art competing methods in both anomaly detection and clustering tasks. Unsupervised representation learning aims at automatically extracting expressive feature representations from data without any manually labelled data. Due to the remarkable capability to learn semantic-rich features, deep neural networks have been becoming one widely-used technique to empower a broad range of machine learning tasks. One main issue with these deep learning techniques is that a massive amount of labelled data is typically required to successfully learn these expressive features. As a result, their transformation power is largely reduced for tasks that are unsupervised in nature, such as anomaly detection and clustering. This is also true to critical domains, such as healthcare and fintech, where collecting massive labelled data is prohibitively expensive and/or is impossible to scale. To bridge this gap, in this work we explore fully unsupervised representation learning techniques to enable downstream unsupervised learning methods on those critical domains. In recent years, many unsupervised representation learning methods (Mikolov et al., 2013a; Le & Mikolov, 2014; Misra et al., 2016; Lee et al., 2017; Gidaris et al., 2018) have been introduced, of which most are self-supervised approaches that formulate the problem as an annotation free pretext task.

This paper describes a novel deep learning-based method for mitigating the effects of atmospheric distortion. We have built an end-to-end supervised convolutional neural network (CNN) to reconstruct turbulence-corrupted video sequence. Our framework has been developed on the residual learning concept, where the spatio-temporal distortions are learnt and predicted. Our experiments demonstrate that the proposed method can deblur, remove ripple effect and enhance contrast of the video sequences simultaneously. Our model was trained and tested with both simulated and real distortions. Experimental results of the real distortions show that our method outperforms the existing ones by up to 3.8% in term of the quality of restored images, and it achieves faster speed than the state-of-the-art methods by up to 23 times with GPU implementation.

The customizable nature of deep learning models have allowed them to be successful predictors in various disciplines. These models are often trained with respect to thousands or millions of instances for complicated problems, but the gathering of such an immense collection may be infeasible and expensive. However, what often occurs is the pollution of redundant information from these instances to the deep learning models. This paper outlines an algorithm that dynamically selects and appends instances to a training dataset from uncertain regions of the parameter space based on differences in predictions from multiple convolutional neural networks (CNNs). These CNNs are also simultaneously trained on this growing dataset to construct more accurate and knowledgable models. The methodology presented has reduced training dataset sizes by almost 90% and maintained predictive power in two diagnostics of high energy density physics.

Predictive process analytics has recently gained significant attention, and yet its successful adoption in organisations relies on how well users can trust the predictions of the underlying machine learning algorithms that are often applied and recognised as a `black-box'. Without understanding the rationale of the black-box machinery, there will be a lack of trust in the predictions, a reluctance to use the predictions, and in the worse case, consequences of an incorrect decision based on the prediction. In this paper, we emphasise the importance of interpreting the predictive models in addition to the evaluation using conventional metrics, such as accuracy, in the context of predictive process monitoring. We review existing studies on business process monitoring benchmarks for predicting process outcomes and remaining time. We derive explanations that present the behaviour of the entire predictive model as well as explanations describing a particular prediction. These explanations are used to reveal data leakages, assess the interpretability of features used by the model, and the degree of the use of process knowledge in the existing benchmark models. Findings from this exploratory study motivate the need to incorporate interpretability in predictive process analytics.

--We introduce HT AD, a novel model for diagnosis prediction using Electronic Health Records (EHR) represented as Heterogeneous Information Networks. Recent studies on modeling EHR have shown success in automatically learning representations of the clinical records in order to avoid the need for manual feature selection. However, these representations are often learned and aggregated without specificity for the different possible targets being predicted. Our model introduces a target-aware hierarchical attention mechanism that allows it to learn to attend to the most important clinical records when aggregating their representations for prediction of a diagnosis. We evaluate our model using a publicly available benchmark dataset and demonstrate that the use of target-aware attention significantly improves performance compared to the current state of the art. Additionally, we propose a method for incorporating non-categorical data into our predictions and demonstrate that this technique leads to further performance improvements. Lastly, we demonstrate that the predictions made by our proposed model are easily interpretable. I NTRODUCTION Electronic Health Records (EHR) provide a comprehensive picture of patients' medical histories, consisting of information such as written clinician notes, medical imagery, prescriptions, and diagnoses.

Background: Three-dimensional, whole heart, balanced steady state free precession (WH-bSSFP) sequences provide delineation of intra-cardiac and vascular anatomy. However, they have long acquisition times. Here, we propose significant speed ups using a deep learning single volume super resolution reconstruction, to recover high resolution features from rapidly acquired low resolution WH-bSSFP images. Methods: A 3D residual U-Net was trained using synthetic data, created from a library of high-resolution WH-bSSFP images by simulating 0.5 slice resolution and 0.5 phase resolution. The trained network was validated with synthetic test data, as well as prospective low-resolution data. Results: Synthetic low-resolution data had significantly better image quality after super-resolution reconstruction. Qualitative image scores showed super-resolved images had better edge sharpness, fewer residual artefacts and less image distortion than low-resolution images, with similar scores to high-resolution data. Quantitative image scores showed super-resolved images had significantly better edge sharpness than low-resolution or high-resolution images, with significantly better signal-to-noise ratio than high-resolution data. Vessel diameters measurements showed over-estimation in the low-resolution measurements, compared to the high-resolution data. No significant differences and no bias was found in the super-resolution measurements. Conclusion: This paper demonstrates the potential of using a residual U-Net for super-resolution reconstruction of rapidly acquired low-resolution whole heart bSSFP data within a clinical setting. The resulting network can be applied very quickly, making these techniques particularly appealing within busy clinical workflow. Thus, we believe that this technique may help speed up whole heart CMR in clinical practice.

Evaluation of Bayesian deep learning (BDL) methods is challenging. We often seek to evaluate the methods' robustness and scalability, assessing whether new tools give `better' uncertainty estimates than old ones. These evaluations are paramount for practitioners when choosing BDL tools on-top of which they build their applications. Current popular evaluations of BDL methods, such as the UCI experiments, are lacking: Methods that excel with these experiments often fail when used in application such as medical or automotive, suggesting a pertinent need for new benchmarks in the field. We propose a new BDL benchmark with a diverse set of tasks, inspired by a real-world medical imaging application on \emph{diabetic retinopathy diagnosis}. Visual inputs (512x512 RGB images of retinas) are considered, where model uncertainty is used for medical pre-screening---i.e. to refer patients to an expert when model diagnosis is uncertain. Methods are then ranked according to metrics derived from expert-domain to reflect real-world use of model uncertainty in automated diagnosis. We develop multiple tasks that fall under this application, including out-of-distribution detection and robustness to distribution shift. We then perform a systematic comparison of well-tuned BDL techniques on the various tasks. From our comparison we conclude that some current techniques which solve benchmarks such as UCI `overfit' their uncertainty to the dataset---when evaluated on our benchmark these underperform in comparison to simpler baselines. The code for the benchmark, its baselines, and a simple API for evaluating new BDL tools are made available at https://github.com/oatml/bdl-benchmarks.

With the growth of using cell phones and the increase in diversity of smart mobile devices, a massive volume of data is generated continuously in the process of using these devices. Among these data, Call Detail Records, CDR, is highly remarkable. Since CDR contains both temporal and spatial labels, mobility analysis of CDR is one of the favorite subjects of study among the researchers. The user next location prediction is one of the main problems in the field of human mobility analysis. In this paper, we propose a data processing framework to predict user next location. We propose domain-specific data processing strategies and design a deep neural network model which is based on recurrent neurons and perform regression tasks. Using this prediction framework, the error of the prediction decreases from 74% to 55% in comparison to the worst and best performing traditional models. Methods, strategies, the framework and the results of this paper can be helpful in many applications such as urban planning and digital marketing.

Australia Abstract Stability of recurrent models is closely linked with trainability, generalizability and in some applications, safety. Methods that train stable recurrent neural networks, however, do so at a significant cost to expressibility. We propose an implicit model structure that allows for a convex parametriza-tion of stable models using contraction analysis of nonlinear systems. Using these stability conditions we propose a new approach to model initialization and then provide a number of empirical results comparing the performance of our proposed model set to previous stable RNNs and vanilla RNNs. By carefully controlling stability in the model, we observe a significant increase in the speed of training and model performance. Keywords: System Identification, Contraction, Stability, Recurrent Neural Network, V anishing Gradient, Exploding Gradient, Nonlinear Systems, Echo State Network Notation Most of our notation is standard.