Visual aesthetics has been shown to critically affect a variety of constructs such as perceived usability, satisfaction, and pleasure. However, visual aesthetics is also a subjective concept and therefore, presents its unique challenges in training a machine learning algorithm to learn such subjectiveness. Given the importance of visual aesthetics in human-computer interaction, it is vital that machines adequately assess the concept of visual aesthetics. Machine learning, especially deep learning techniques have already shown great promise on tasks with well-defined goals such as identifying objects in images or translating from one language to another. However, quantification of image aesthetics has been one of the most persistent problems in image processing and computer vision.

End of life care might be improved with Deep Learning. An AI program in a successful pilot study predicted how long people will live. George Dvorsky in Gizmodo and others reported on their work. The Stanford University team is using an algorithm to predict mortality, and their goal is to improve timing of end-of-life care for critically ill patients. While 80 percent of Americans prefer to spend their final days in their home, only 20 percent do just that.

When it comes to test how correct the gradient in deep neural nets, how can we check? I have found good solution though, couldn't understand why it requires us to calculate the estimate gradient twice.

Theoretical understanding of how deep neural network (DNN) extracts features from input images is still unclear, but it is widely believed that the extraction is performed hierarchically through a process of coarse-graining. It reminds us of the basic concept of renormalization group (RG) in statistical physics. In order to explore possible relations between DNN and RG, we use the Restricted Boltzmann machine (RBM) applied to Ising model and construct a flow of model parameters (in particular, temperature) generated by the RBM. We show that the unsupervised RBM trained by spin configurations at various temperatures from $T=0$ to $T=6$ generates a flow along which the temperature approaches the critical value $T_c=2.27$. This behavior is opposite to the typical RG flow of the Ising model. By analyzing various properties of the weight matrices of the trained RBM, we discuss why it flows towards $T_c$ and how the RBM learns to extract features of spin configurations.

Activation functions have a notorious impact on neural networks on both training and testing the models against the desired problem. Currently, the most used activation function is the Rectified Linear Unit (ReLU). This paper introduces a new and novel activation function, closely related with the new activation $Swish = x * sigmoid(x)$ (Ramachandran et al., 2017) which generalizes it. We call the new activation $E-swish = \beta x * sigmoid(x)$. We show that E-swish outperforms many other well-known activations including both ReLU and Swish. For example, using E-swish provided 1.5% and 4.6% accuracy improvements on Cifar10 and Cifar100 respectively for the WRN 10-2 when compared to ReLU and 0.35% and 0.6% respectively when compared to Swish. The code to reproduce all our experiments can be found at https://github.com/EricAlcaide/E-swish

This work is an endeavor to develop a deep learning methodology for automated anatomical labeling of a given region of interest (ROI) in brain computed tomography (CT) scans. We combine both local and global context to obtain a representation of the ROI. We then use Relation Networks (RNs) to predict the corresponding anatomy of the ROI based on its relationship score for each class. Further, we propose a novel strategy employing nearest neighbors approach for training RNs. We train RNs to learn the relationship of the target ROI with the joint representation of its nearest neighbors in each class instead of all data-points in each class. The proposed strategy leads to better training of RNs along with increased performance as compared to training baseline RN network.

Labeled data used for training activity recognition classifiers are usually limited in terms of size and diversity. Thus, the learned model may not generalize well when used in real-world use cases. Semi-supervised learning augments labeled examples with unlabeled examples, often resulting in improved performance. However, the semi-supervised methods studied in the activity recognition literatures assume that feature engineering is already done. In this paper, we lift this assumption and present two semi-supervised methods based on convolutional neural networks (CNNs) to learn discriminative hidden features. Our semi-supervised CNNs learn from both labeled and unlabeled data while also performing feature learning on raw sensor data. In experiments on three real world datasets, we show that our CNNs outperform supervised methods and traditional semi-supervised learning methods by up to 18% in mean F1-score (Fm).

Many interesting problems in machine learning are being revisited with new deep learning tools. For graph-based semisupervised learning, a recent important development is graph convolutional networks (GCNs), which nicely integrate local vertex features and graph topology in the convolutional layers. Although the GCN model compares favorably with other state-of-the-art methods, its mechanisms are not clear and it still requires a considerable amount of labeled data for validation and model selection. In this paper, we develop deeper insights into the GCN model and address its fundamental limits. First, we show that the graph convolution of the GCN model is actually a special form of Laplacian smoothing, which is the key reason why GCNs work, but it also brings potential concerns of over-smoothing with many convolutional layers. Second, to overcome the limits of the GCN model with shallow architectures, we propose both co-training and self-training approaches to train GCNs. Our approaches significantly improve GCNs in learning with very few labels, and exempt them from requiring additional labels for validation. Extensive experiments on benchmarks have verified our theory and proposals.

The recent success of Deep Neural Networks (DNNs) has drastically improved the state of the art for many application domains. While achieving high accuracy performance, deploying state-of-the-art DNNs is a challenge since they typically require billions of expensive arithmetic computations. In addition, DNNs are typically deployed in ensemble to boost accuracy performance, which further exacerbates the system requirements. This computational overhead is an issue for many platforms, e.g. data centers and embedded systems, with tight latency and energy budgets. In this article, we introduce flexible DNNs ensemble processing technique, which achieves large reduction in average inference latency while incurring small to negligible accuracy drop. Our technique is flexible in that it allows for dynamic adaptation between quality of results (QoR) and execution runtime. We demonstrate the effectiveness of the technique on AlexNet and ResNet-50 using the ImageNet dataset. This technique can also easily handle other types of networks.

The field of machine learning offers many potent models for inference. Unfortunately, simply optimizing how well these models perform on a fixed training sample often leads to relatively poor performance on new test data compared to models that fit the training data less well. Regularization schemes are used to constrain the fitted model to improve performance on new data. One popular regularization tactic is to corrupt the training data with independently sampled noise. This constrains the model to work on data that is different from the original training data in a way that does not change the correct inference.