A deep learning classifier for detecting seizures in neonates is proposed. This architecture is designed to detect seizure events from raw electroencephalogram (EEG) signals as opposed to the state-of-the-art hand engineered feature-based representation employed in traditional machine learning based solutions. The seizure detection system utilises only convolutional layers in order to process the multichannel time domain signal and is designed to exploit the large amount of weakly labelled data in the training stage. The system performance is assessed on a large database of continuous EEG recordings of 834h in duration; this is further validated on a held-out publicly available dataset and compared with two baseline SVM based systems. The developed system achieves a 56% relative improvement with respect to a feature-based state-of-the art baseline, reaching an AUC of 98.5%; this also compares favourably both in terms of performance and run-time. The effect of varying architectural parameters is thoroughly studied. The performance improvement is achieved through novel architecture design which allows more efficient usage of available training data and end-to-end optimisation from the front-end feature extraction to the back-end classification. The proposed architecture opens new avenues for the application of deep learning to neonatal EEG, where the performance becomes a function of the amount of training data with less dependency on the availability of precise clinical labels.

Many existing conversation models that are based on the encoder-decoder framework have focused on ways to make the encoder more complicated to enrich the context vectors so as to increase the diversity and informativeness of generated responses. However, these approaches face two problems. First, the decoder is too simple to effectively utilize the previously generated information and tends to generate duplicated and self-contradicting responses. Second, the complex encoder tends to generate diverse but incoherent responses because the complex context vectors may deviate from the original semantics of context. In this work, we proposed a conversation model named "THINK" (Teamwork generation Hover around Impressive Noticeable Keywords) to make the decoder more complicated and avoid generating duplicated and self-contradicting responses. The model simplifies the context vectors and increases the coherence of generated responses in a reasonable way. For this model, we propose Teamwork generation framework and Semantics Extractor. Compared with other baselines, both automatic and human evaluation showed the advantages of our model.

Travel time estimation is a basic but important part in intelligent transportation systems, especially widely applied in online map services to help travel navigation and route planning. Most previous works commonly model the road segments or intersections separately and obtain their spatial-temporal characteristics for travel time estimation. However, due to the continuous alternation of the road segments and intersections, the dynamic features of them are supposed to be coupled and interactive. Therefore, modeling one of them limits further improvement in accuracy of estimating travel time. To address the above problems, we propose a novel graph-based deep learning framework for travel time estimation, namely Spatial-Temporal Dual Graph Neural Networks (STDGNN). Specifically, we first establish the spatial-temporal dual graph architecture to capture the complex correlations of both intersections and road segments. The adjacency relations of intersections and that of road segments are respectively characterized by node-wise graph and edge-wise graph. In order to capture the joint spatial-temporal dynamics of the intersections and road segments, we adopt the spatial-temporal learning layer that incorporates the multi-scale spatial-temporal graph convolution networks and dual graph interaction networks. Followed by the spatial-temporal learning layer, we also employ the multi-task learning layer to estimate the travel time of a given whole route and each road segment simultaneously. We conduct extensive experiments to evaluate our proposed model on two real-world trajectory datasets, and the experimental results show that STDGNN significantly outperforms several state-of-art baselines.

We present a new class of adaptive stochastic optimization algorithms, which overcomes many of the known shortcomings of popular adaptive optimizers that are currently used for the fine tuning of artificial neural networks (ANNs). Its underpinning theory relies on advances of Euler's polygonal approximations for stochastic differential equations (SDEs) with monotone coefficients. As a result, it inherits the stability properties of tamed algorithms, while it addresses other known issues, e.g. vanishing gradients in ANNs. In particular, we provide an nonasymptotic analysis and full theoretical guarantees for the convergence properties of an algorithm of this novel class, which we named TH$\varepsilon$O POULA (or, simply, TheoPouLa). Finally, several experiments are presented with different types of ANNs, which show the superior performance of TheoPouLa over many popular adaptive optimization algorithms.

Anomalies represent deviations from the intended system operation and can lead to decreased efficiency as well as partial or complete system failure. As the causes of anomalies are often unknown due to complex system dynamics, efficient anomaly detection is necessary. Conventional detection approaches rely on statistical and time-invariant methods that fail to address the complex and dynamic nature of anomalies. With advances in artificial intelligence and increasing importance for anomaly detection and prevention in various domains, artificial neural network approaches enable the detection of more complex anomaly types while considering temporal and contextual characteristics. In this article, a survey on state-of-the-art anomaly detection using deep neural and especially long short-term memory networks is conducted. The investigated approaches are evaluated based on the application scenario, data and anomaly types as well as further metrics. To highlight the potential of upcoming anomaly detection techniques, graph-based and transfer learning approaches are also included in the survey, enabling the analysis of heterogeneous data as well as compensating for its shortage and improving the handling of dynamic processes.

Momentum strategies are an important part of alternative investments and are at the heart of commodity trading advisors (CTAs). These strategies have however been found to have difficulties adjusting to rapid changes in market conditions, such as during the 2020 market crash. In particular, immediately after momentum turning points, where a trend reverses from an uptrend (downtrend) to a downtrend (uptrend), time-series momentum (TSMOM) strategies are prone to making bad bets. To improve the response to regime change, we introduce a novel approach, where we insert an online change-point detection (CPD) module into a Deep Momentum Network (DMN) [1904.04912] pipeline, which uses an LSTM deep-learning architecture to simultaneously learn both trend estimation and position sizing. Furthermore, our model is able to optimise the way in which it balances 1) a slow momentum strategy which exploits persisting trends, but does not overreact to localised price moves, and 2) a fast mean-reversion strategy regime by quickly flipping its position, then swapping it back again to exploit localised price moves. Our CPD module outputs a changepoint location and severity score, allowing our model to learn to respond to varying degrees of disequilibrium, or smaller and more localised changepoints, in a data driven manner. Using a portfolio of 50, liquid, continuous futures contracts over the period 1990-2020, the addition of the CPD module leads to an improvement in Sharpe ratio of $33\%$. Even more notably, this module is especially beneficial in periods of significant nonstationarity, and in particular, over the most recent years tested (2015-2020) the performance boost is approximately $400\%$. This is especially interesting as traditional momentum strategies have been underperforming in this period.

Nowadays, the videos on the Internet are prevailing. The precise and in-depth understanding of the videos is a difficult but valuable problem for both platforms and researchers. The existing video understand models do well in object recognition tasks but currently still cannot understand the abstract and contextual features like highlight humor frames in comedy videos. The current industrial works are also mainly focused on the basic category classification task based on the appearances of objects. The feature detection methods for the abstract category remains blank. A data structure that includes the information of video frames, audio spectrum and texts provide a new direction to explore. The multimodal models are proposed to make this in-depth video understanding mission possible. In this paper, we analyze the difficulties in abstract understanding of videos and propose a multimodal structure to obtain state-of-the-art performance in this field. Then we select several benchmarks for multimodal video understanding and apply the most suitable model to find the best performance. At last, we evaluate the overall spotlights and drawbacks of the models and methods in this paper and point out the possible directions for further improvements.

Recent years have seen the introduction of a range of methods for post-hoc explainability of image classifier predictions. However, these post-hoc explanations may not always align perfectly with classifier predictions, which poses a significant challenge when attempting to debug models based on such explanations. To this end, we seek a methodology that can improve alignment between model predictions and explanation method that is both agnostic to the model and explanation classes and which does not require ground truth explanations. We achieve this through a novel explanation-driven data augmentation (EDDA) method that augments the training data with occlusions of existing data stemming from model-explanations; this is based on the simple motivating principle that occluding salient regions for the model prediction should decrease the model confidence in the prediction, while occluding non-salient regions should not change the prediction -- if the model and explainer are aligned. To verify that this augmentation method improves model and explainer alignment, we evaluate the methodology on a variety of datasets, image classification models, and explanation methods. We verify in all cases that our explanation-driven data augmentation method improves alignment of the model and explanation in comparison to no data augmentation and non-explanation driven data augmentation methods. In conclusion, this approach provides a novel model- and explainer-agnostic methodology for improving alignment between model predictions and explanations, which we see as a critical step forward for practical deployment and debugging of image classification models.

In this paper, we study the task of source-free domain adaptation (SFDA), where the source data are not available during target adaptation. Previous works on SFDA mainly focus on aligning the cross-domain distributions. However, they ignore the generalization ability of the pretrained source model, which largely influences the initial target outputs that are vital to the target adaptation stage. To address this, we make the interesting observation that the model accuracy is highly correlated with whether or not attention is focused on the objects in an image. To this end, we propose a generic and effective framework based on Transformer, named TransDA, for learning a generalized model for SFDA. Specifically, we apply the Transformer as the attention module and inject it into a convolutional network. By doing so, the model is encouraged to turn attention towards the object regions, which can effectively improve the model's generalization ability on the target domains. Moreover, a novel self-supervised knowledge distillation approach is proposed to adapt the Transformer with target pseudo-labels, thus further encouraging the network to focus on the object regions. Experiments on three domain adaptation tasks, including closed-set, partial-set, and open-set adaption, demonstrate that TransDA can greatly improve the adaptation accuracy and produce state-of-the-art results. The source code and trained models are available at https://github.com/ygjwd12345/TransDA.

In a real-world setting biological agents do not have infinite resources to learn new things. It is thus useful to recycle previously acquired knowledge in a way that allows for faster, less resource-intensive acquisition of multiple new skills. Neural networks in the brain are likely not entirely re-trained with new tasks, but how they leverage existing computations to learn new tasks is not well understood. In this work, we study this question in artificial neural networks trained on commonly used neuroscience paradigms. Building on recent work from the multi-task learning literature, we propose two ingredients: (1) network modularity, and (2) learning task primitives. Together, these ingredients form inductive biases we call structural and functional, respectively. Using a corpus of nine different tasks, we show that a modular network endowed with task primitives allows for learning multiple tasks well while keeping parameter counts, and updates, low. We also show that the skills acquired with our approach are more robust to a broad range of perturbations compared to those acquired with other multi-task learning strategies. This work offers a new perspective on achieving efficient multi-task learning in the brain, and makes predictions for novel neuroscience experiments in which targeted perturbations are employed to explore solution spaces.