In this paper, we exploit the benefits of the deep learning approach to design an efficient system of Amazigh handwritten recognition. Indeed, this approach has proved a greater efficiency in the various domains, especially recognition tasks. However, to take full advantage of this approach it's necessary to construct an adequate dataset of training and testing that represent faithfully the concerned problem. To this end, we have prepared our dataset of 102 writers each one contains 33 characters of IRCAM-Tifinagh. Inspired by the MNIST database, the set of characters is size-normalized and centered in a fixed-size image. The resulting is a grey level image of size 28x28, where the black color is the non-color of the character. The number of images produced after this preprocessing step is 3,366.
Heart failure hospitalization is a severe burden on healthcare. How to predict and therefore prevent readmission has been a significant challenge in outcomes research. To address this, we propose a deep learning approach to predict readmission from clinical notes. Unlike conventional methods that use structured data for prediction, we leverage the unstructured clinical notes to train deep learning models based on convolutional neural networks (CNN). We then use the trained models to classify and predict potentially high-risk admissions/patients. For evaluation, we trained CNNs using the discharge summary notes in the MIMIC III database. We also trained regular machine learning models based on random forest using the same datasets. The result shows that deep learning models outperform the regular models in prediction tasks. CNN method achieves a F1 score of 0.756 in general readmission prediction and 0.733 in 30-day readmission prediction, while random forest only achieves a F1 score of 0.674 and 0.656 respectively. We also propose a chi-square test based method to interpret key features associated with deep learning predicted readmissions. It reveals clinical insights about readmission embedded in the clinical notes. Collectively, our method can make the human evaluation process more efficient and potentially facilitate the reduction of readmission rates.
Current supervised deep learning frameworks rely on annotated data for modeling the underlying data distribution of a given task. In particular for computer vision algorithms powered by deep learning, the quality of annotated data is the most critical factor in achieving the desired algorithm performance. Data annotation is, typically, a manual process where the annotator follows guidelines and operates in a best-guess manner. Labeling criteria among annotators can show discrepancies in labeling results. This may impact the algorithm inference performance. Given the popularity and widespread use of deep learning among computer vision, more and more custom datasets are needed to train neural networks to tackle different kinds of tasks. Unfortunately, there is no full understanding of the factors that affect annotated data quality, and how it translates into algorithm performance. In this paper we studied this problem for object detection and recognition.We conducted several data annotation experiments to measure inter annotator agreement and consistency, as well as how the selection of ground truth impacts the perceived algorithm performance.We propose a methodology to monitor the quality of annotations during the labeling of images and how it can be used to measure performance. We also show that neglecting to monitor the annotation process can result in significant loss in algorithm precision. Through these experiments, we observe that knowledge of the labeling process, training data, and ground truth data used for algorithm evaluation are fundamental components to accurately assess trustworthiness of an AI system.
We build on the dynamical systems approach to deep learning, where deep residual networks are idealized as continuous-time dynamical systems, from the approximation perspective. In particular, we establish general sufficient conditions for universal approximation using continuous-time deep residual networks, which can also be understood as approximation theories in $L^p$ using flow maps of dynamical systems. In specific cases, rates of approximation in terms of the time horizon are also established. Overall, these results reveal that composition function approximation through flow maps present a new paradigm in approximation theory and contributes to building a useful mathematical framework to investigate deep learning.
This paper examines the applicability in realistic scenarios of two deep learning based solutions to the overlapping speaker separation problem. Firstly, we present experiments that show that these methods are applicable for a broad range of languages. Further experimentation indicates limited performance loss for untrained languages, when these have common features with the trained language(s). Secondly, it investigates how the methods deal with realistic background noise and proposes some modifications to better cope with these disturbances. The deep learning methods that will be examined are deep clustering and deep attractor networks.
With the advent of the era of big data, deep learning has become a prevalent building block in a variety of machine learning or data mining tasks, such as signal processing, network modeling and traffic analysis, to name a few. The massive user data crowdsourced plays a crucial role in the success of deep learning models. However, it has been shown that user data may be inferred from trained neural models and thereby exposed to potential adversaries, which raises information security and privacy concerns. To address this issue, recent studies leverage the technique of differential privacy to design private-preserving deep learning algorithms. Albeit successful at privacy protection, differential privacy degrades the performance of neural models. In this paper, we develop ADADP, an adaptive and fast convergent learning algorithm with a provable privacy guarantee. ADADP significantly reduces the privacy cost by improving the convergence speed with an adaptive learning rate and mitigates the negative effect of differential privacy upon the model accuracy by introducing adaptive noise. The performance of ADADP is evaluated on real-world datasets. Experiment results show that it outperforms state-of-the-art differentially private approaches in terms of both privacy cost and model accuracy.
Deep learning demands a huge amount of well-labeled data to train the network parameters. How to use the least amount of labeled data to obtain the desired classification accuracy is of great practical significance, because for many real-world applications (such as medical diagnosis), it is difficult to obtain so many labeled samples. In this paper, modify the unsupervised discriminant projection algorithm from dimension reduction and apply it as a regularization term to propose a new semi-supervised deep learning algorithm, which is able to utilize both the local and nonlocal distribution of abundant unlabeled samples to improve classification performance. Experiments show that given dozens of labeled samples, the proposed algorithm can train a deep network to attain satisfactory classification results.
We study a deep learning (DL) based limited feedback methods for multi-antenna systems. Deep neural networks (DNNs) are introduced to replace an end-to-end limited feedback procedure including pilot-aided channel training process, channel codebook design, and beamforming vector selection. The DNNs are trained to yield binary feedback information as well as an efficient beamforming vector which maximizes the effective channel gain. Compared to conventional limited feedback schemes, the proposed DL method shows an 1 dB symbol error rate (SER) gain with reduced computational complexity.
When and why can a neural network be successfully trained? This article provides an overview of optimization algorithms and theory for training neural networks. First, we discuss the issue of gradient explosion/vanishing and the more general issue of undesirable spectrum, and then discuss practical solutions including careful initialization and normalization methods. Second, we review generic optimization methods used in training neural networks, such as SGD, adaptive gradient methods and distributed methods, and theoretical results for these algorithms. Third, we review existing research on the global issues of neural network training, including results on bad local minima, mode connectivity, lottery ticket hypothesis and infinite-width analysis.
Semi-Supervised Learning (SSL) algorithms have shown great potential in training regimes when access to labeled data is scarce but access to unlabeled data is plentiful. However, our experiments illustrate several shortcomings that prior SSL algorithms suffer from. In particular, poor performance when unlabeled and labeled data distributions differ. To address these observations, we develop RealMix, which achieves state-of-the-art results on standard benchmark datasets across different labeled and unlabeled set sizes while overcoming the aforementioned challenges. Notably, RealMix achieves an error rate of 9.79% on CIFAR10 with 250 labels and is the only SSL method tested able to surpass baseline performance when there is significant mismatch in the labeled and unlabeled data distributions. RealMix demonstrates how SSL can be used in real world situations with limited access to both data and compute and guides further research in SSL with practical applicability in mind.
Automated recognition and classification of bacteria species from microscopic images have significant importance in clinical microbiology. Bacteria classification is usually carried out manually by biologists using different shapes and morphologic characteristics of bacteria species. The manual taxonomy of bacteria types from microscopy images is time-consuming and a challenging task for even experienced biologists. In this study, an automated deep learning based classification approach has been proposed to classify bacterial images into different categories. The ResNet-50 pre-trained CNN architecture has been used to classify digital bacteria images into 33 categories. The transfer learning technique was employed to accelerate the training process of the network and improve the classification performance of the network. The proposed method achieved an average classification accuracy of 99.2%. The experimental results demonstrate that the proposed technique surpasses state-of-the-art methods in the literature and can be used for any type of bacteria classification tasks.
We propose a novel quaternion product unit (QPU) to represent data on 3D rotation groups. The QPU leverages quaternion algebra and the law of 3D rotation group, representing 3D rotation data as quaternions and merging them via a weighted chain of Hamilton products. We prove that the representations derived by the proposed QPU can be disentangled into "rotation-invariant" features and "rotation-equivariant" features, respectively, which supports the rationality and the efficiency of the QPU in theory. We design quaternion neural networks based on our QPUs and make our models compatible with existing deep learning models. Experiments on both synthetic and real-world data show that the proposed QPU is beneficial for the learning tasks requiring rotation robustness.
In medical science, the use of computer science in disease detection and diagnosis is gaining popularity. Previously, the detection of disease used to take a significant amount of time and was less reliable. Machine learning (ML) techniques employed in recent biomedical researches are making revolutionary changes by gaining higher accuracy with more concise timing. At present, it is even possible to automatically detect diseases from the scanned images with the help of ML. In this research, we have taken such an attempt to detect retinal diseases from optical coherence tomography (OCT) X-ray images. Here, we propose a deep learning (DL) based approach in detecting retinal diseases from OCT images which can identify three conditions of the retina. Four different models used in this approach are compared with each other. On the test set, the detection accuracy is 98.00\% for a vanilla convolutional neural network (CNN) model, 99.07\% for Xception model, 97.00\% for ResNet50 model, and 99.17\% for MobileNetV2 model. The MobileNetV2 model acquires the highest accuracy, and the closest to the highest is the Xception model. The proposed approach has a potential impact on creating a tool for automatically detecting retinal diseases.
Myocardial infarction is the leading cause of death worldwide. In this paper, we design domain-inspired neural network models to detect myocardial infarction. First, we study the contribution of various leads. This systematic analysis, first of its kind in the literature, indicates that out of 15 ECG leads, data from the v6, vz, and ii leads are critical to correctly identify myocardial infarction. Second, we use this finding and adapt the ConvNetQuake neural network model--originally designed to identify earthquakes--to attain state-of-the-art classification results for myocardial infarction, achieving $99.43\%$ classification accuracy on a record-wise split, and $97.83\%$ classification accuracy on a patient-wise split. These two results represent cardiologist-level performance level for myocardial infarction detection after feeding only 10 seconds of raw ECG data into our model. Third, we show that our multi-ECG-channel neural network achieves cardiologist-level performance without the need of any kind of manual feature extraction or data pre-processing.
Total column water vapor is an important factor for the weather and climate. This study apply deep learning based multiple regression to map the TCWV with elements that can improve spatiotemporal prediction. In this study, we predict the TCWV with the use of ERA5 that is the fifth generation ECMWF atmospheric reanalysis of the global climate. We use an appropriate deep learning based multiple regression algorithm using Keras library to improve nonlinear prediction between Total Column water vapor and predictors as Mean sea level pressure, Surface pressure, Sea surface temperature, 100 metre U wind component, 100 metre V wind component, 10 metre U wind component, 10 metre V wind component, 2 metre dew point temperature, 2 metre temperature. The results obtained permit to build a predictor which modelling TCWV with a mean abs error (MAE) equal to 3.60 kg/m2 and a coefficient of determination R2 equal to 0.90.
Accurate automatic optimization heuristics are necessary for dealing with the complexity and diversity of modern hardware and software. Machine learning is a proven technique for learning such heuristics, but its success is bound by the quality of the features used. These features must be hand crafted by developers through a combination of expert domain knowledge and trial and error. This makes the quality of the final model directly dependent on the skill and available time of the system architect. Our work introduces a better way for building heuristics. We develop a deep neural network that learns heuristics over raw code, entirely without using code features. The neural network simultaneously constructs appropriate representations of the code and learns how best to optimize, removing the need for manual feature creation. Further, we show that our neural nets can transfer learning from one optimization problem to another, improving the accuracy of new models, without the help of human experts. We compare the effectiveness of our automatically generated heuristics against ones with features hand-picked by experts. We examine two challenging tasks:predicting optimal mapping for heterogeneous parallelism and GPU thread coarsening factors. In 89% of the cases, the quality of our fully automatic heuristics matches or surpasses that of state-of-the-art predictive models using hand-crafted features, providing on average 14% and 12% more performance with no human effort expended on designing features.
Differentiable programming is the combination of classical neural networks modules with algorithmic ones in an end-to-end differentiable model. These new models, that use automatic differentiation to calculate gradients, have new learning capabilities (reasoning, attention and memory). In this tutorial, aimed at researchers in nonlinear systems with prior knowledge of deep learning, we present this new programming paradigm, describe some of its new features such as attention mechanisms, and highlight the benefits they bring. Then, we analyse the uses and limitations of traditional deep learning models in the modeling and prediction of dynamical systems. Here, a dynamical system is meant to be a set of state variables that evolve in time under general internal and external interactions. Finally, we review the advantages and applications of differentiable programming to dynamical systems.
Currently there are several well-known approaches to non-intrusive appliance load monitoring rule based, stochastic finite state machines, neural networks and sparse coding. Recently several studies have proposed a new approach based on multi label classification. Different appliances are treated as separate classes, and the task is to identify the classes given the aggregate smart-meter reading. Prior studies in this area have used off the shelf algorithms like MLKNN and RAKEL to address this problem. In this work, we propose a deep learning based technique. There are hardly any studies in deep learning based multi label classification; two new deep learning techniques to solve the said problem are fundamental contributions of this work. These are deep dictionary learning and deep transform learning. Thorough experimental results on benchmark datasets show marked improvement over existing studies.
VLSI mask optimization is one of the most critical stages in manufacturability aware design, which is costly due to the complicated mask optimization and lithography simulation. Recent researches have shown prominent advantages of machine learning techniques dealing with complicated and big data problems, which bring potential of dedicated machine learning solution for DFM problems and facilitate the VLSI design cycle. In this paper, we focus on a heterogeneous OPC framework that assists mask layout optimization. Preliminary results show the efficiency and effectiveness of proposed frameworks that have the potential to be alternatives to existing EDA solutions.
Computer vision has shown promising results in medical image processing. Pneumothorax is a deadly condition and if not diagnosed and treated at time then it causes death. It can be diagnosed with chest X-ray images. We need an expert and experienced radiologist to predict whether a person is suffering from pneumothorax or not by looking at the chest X-ray images. Everyone does not have access to such a facility. Moreover, in some cases, we need quick diagnoses. So we propose an image segmentation model to predict and give the output a mask that will assist the doctor in taking this crucial decision. Deep Learning has proved their worth in many areas and outperformed man state-of-the-art models. We want to use the power of these deep learning model to solve this problem. We have used U-net [13] architecture with ResNet [17] as a backbone and achieved promising results. U-net [13] performs very well in medical image processing and semantic segmentation. Our problem falls in the semantic segmentation category.