Neural networks have proven to be remarkably successful for a wide range of complicated tasks, from image recognition and object detection to speech recognition and machine translation. One of their successes is the skill in prediction of future dynamics given a suitable training set of data. Previous studies have shown how Echo State Networks (ESNs), a subset of Recurrent Neural Networks, can successfully predict even chaotic systems for times longer than the Lyapunov time. This study shows that, remarkably, ESNs can successfully predict dynamical behavior that is qualitatively different from any behavior contained in the training set. Evidence is provided for a fluid dynamics problem where the flow can transition between laminar (ordered) and turbulent (disordered) regimes. Despite being trained on the turbulent regime only, ESNs are found to predict laminar behavior. Moreover, the statistics of turbulent-to-laminar and laminar-to-turbulent transitions are also predicted successfully, and the utility of ESNs in acting as an early-warning system for transition is discussed. These results are expected to be widely applicable to data-driven modelling of temporal behaviour in a range of physical, climate, biological, ecological and finance models characterized by the presence of tipping points and sudden transitions between several competing states.
Due to the dynamic nature, chaotic time series are difficult predict. In conventional signal processing approaches signals are treated either in time or in space domain only. Spatio-temporal analysis of signal provides more advantages over conventional uni-dimensional approaches by harnessing the information from both the temporal and spatial domains. Herein, we propose an spatio-temporal extension of RBF neural networks for the prediction of chaotic time series. The proposed algorithm utilizes the concept of time-space orthogonality and separately deals with the temporal dynamics and spatial non-linearity(complexity) of the chaotic series. The proposed RBF architecture is explored for the prediction of Mackey-Glass time series and results are compared with the standard RBF. The spatio-temporal RBF is shown to out perform the standard RBFNN by achieving significantly reduced estimation error.
We present Contingency Model Predictive Control (CMPC), a motion planning and control framework that optimizes performance objectives while simultaneously maintaining a contingency plan -- an alternate trajectory that avoids a potential hazard. By preserving the existence of a feasible avoidance trajectory, CMPC anticipates emergency and keeps the controlled system in a safe state that is selectively robust to the identified hazard. We accomplish this by adding an additional prediction horizon in parallel to the typical Model Predictive Control (MPC) horizon. This extra horizon is constrained to guarantee safety from the contingent threat and is coupled to the nominal horizon at its first command. Thus, the two horizons negotiate to compute commands that are both optimized for performance and robust to the contingent event. This article presents a linear formulation for CMPC, illustrates its key features on a toy problem, and then demonstrates its efficacy experimentally on a full-size automated road vehicle that encounters a realistic pop-out obstacle. Contingency MPC approaches potential emergencies with safe, intuitive, and interpretable behavior that balances conservatism with incentive for high performance operation.
Click-through rate prediction plays an important role in the field of recommender system and many other applications. Existing methods mainly extract user interests from user historical behaviors. However, behavioral sequences only contain users' directly interacted items, which are limited by the system's exposure, thus they are often not rich enough to reflect all the potential interests. In this paper, we propose a novel method, named Dynamic Sequential Graph Learning (DSGL), to enhance users or items' representations by utilizing collaborative information from the local sub-graphs associated with users or items. Specifically, we design the Dynamic Sequential Graph (DSG), i.e., a lightweight ego subgraph with timestamps induced from historical interactions. At every scoring moment, we construct DSGs for the target user and the candidate item respectively. Based on the DSGs, we perform graph convolutional operations iteratively in a bottom-up manner to obtain the final representations of the target user and the candidate item. As for the graph convolution, we design a Time-aware Sequential Encoding Layer that leverages the interaction time information as well as temporal dependencies to learn evolutionary user and item dynamics. Besides, we propose a Target-Preference Dual Attention Layer, composed of a preference-aware attention module and a target-aware attention module, to automatically search for parts of behaviors that are relevant to the target and alleviate the noise from unreliable neighbors. Results on real-world CTR prediction benchmarks demonstrate the improvements brought by DSGL.
Automated segmentation of pathological regions of interest has been shown to aid prognosis and follow up treatment. However, accurate pathological segmentations require high quality of annotated data that can be both cost and time intensive to generate. In this work, we propose an automated two-step method that evaluates the quality of medical images from 3D image stacks using a U-net++ model, such that images that can aid further training of the U-net++ model can be detected based on the disagreement in segmentations produced from the final two layers. Images thus detected can then be used to further fine tune the U-net++ model for semantic segmentation. The proposed QU-net++ model isolates around 10\% of images per 3D stack and can scale across imaging modalities to segment cysts in OCT images and ground glass opacity in Lung CT images with Dice cores in the range 0.56-0.72. Thus, the proposed method can be applied for multi-modal binary segmentation of pathology.
In this work, we propose a method to learn probability distributions using sample path data from stochastic differential equations. Specifically, we consider temporally evolving probability distributions (e.g., those produced by integrating local or nonlocal Fokker-Planck equations). We analyze this evolution through machine learning assisted construction of a time-dependent mapping that takes a reference distribution (say, a Gaussian) to each and every instance of our evolving distribution. If the reference distribution is the initial condition of a Fokker-Planck equation, what we learn is the time-T map of the corresponding solution. Specifically, the learned map is a normalizing flow that deforms the support of the reference density to the support of each and every density snapshot in time. We demonstrate that this approach can learn solutions to non-local Fokker-Planck equations, such as those arising in systems driven by both Brownian and L\'evy noise. We present examples with two- and three-dimensional, uni- and multimodal distributions to validate the method.
The paper describes highly-sensitive passive electric potential sensors (EPS) for non-contact detection of multiple biophysical signals, including electrocardiogram (ECG), respiration cycle (RC), and electroencephalogram (EEG). The proposed EPS uses an optimized transimpedance amplifier (TIA), a single guarded sensing electrode, and an adaptive cancellation loop (ACL) to maximize sensitivity (DC transimpedance $=150$~G$\Omega$) in the presence of power line interference (PLI) and motion artifacts. Tests were performed on healthy adult volunteers in noisy and unshielded indoor environments. Useful sensing ranges for ECG, RC, and EEG measurements, as validated against reference contact sensors, were observed to be approximately 50~cm, 100~cm, and 5~cm, respectively. ECG and RC signals were also successfully measured through wooden tables for subjects in sleep-like postures. The EPS were integrated with a wireless microcontroller to realize wireless sensor nodes capable of streaming acquired data to a remote base station in real-time.
This paper presents a method for automatic generation of a training dataset for a deep convolutional neural network used for playing card detection. The solution allows to skip the time-consuming processes of manual image collecting and labelling recognised objects. The YOLOv4 network trained on the generated dataset achieved an efficiency of 99.8% in the cards detection task. The proposed method is a part of a project that aims to automate the process of broadcasting duplicate bridge competitions using a vision system and neural networks.
Recent advances in real-time music score following have made it possible for machines to automatically track highly complex polyphonic music, including full orchestra performances. In this paper, we attempt to take this to an even higher level, namely, live tracking of full operas. We first apply a state-of-the-art audio alignment method based on online Dynamic Time-Warping (OLTW) to full-length recordings of a Mozart opera and, analyzing the tracker's most severe errors, identify three common sources of problems specific to the opera scenario. To address these, we propose a combination of a DTW-based music tracker with specialized audio event detectors (for applause, silence/noise, and speech) that condition the DTW algorithm in a top-down fashion, and show, step by step, how these detectors add robustness to the score follower. However, there remain a number of open problems which we identify as targets for ongoing and future research.
Pre-trained language models in the past years have shown exponential growth in model parameters and compute time. ELECTRA is a novel approach for improving the compute efficiency of pre-trained language models (e.g. BERT) based on masked language modeling (MLM) by addressing the sample inefficiency problem with the replaced token detection (RTD) task. Our work proposes adaptive early exit strategy to maximize the efficiency of the pre-training process by relieving the model's subsequent layers of the need to process latent features by leveraging earlier layer representations. Moreover, we evaluate an initial approach to the problem that has not succeeded in maintaining the accuracy of the model while showing a promising compute efficiency by thoroughly investigating the necessity of the generator module of ELECTRA.