We apply Generative Adversarial Networks (GANs) to the domain of digital pathology. Current machine learning research for digital pathology focuses on diagnosis, but we suggest a different approach and advocate that generative models could help to understand and find fundamental morphological characteristics of cancer tissue. In this paper, we develop a framework which allows GANs to capture key tissue features, and present a vision of how these could link cancer tissue and DNA in the future. To this end, we trained our model on breast cancer tissue from a medium size cohort of 526 patients, producing high fidelity images. We further study how a range of relevant GAN evaluation metrics perform on this task, and propose to evaluate synthetic images with clinically/pathologically meaningful features. Our results show that these models are able to capture key morphological characteristics that link with phenotype, such as survival time and Estrogen-receptor (ER) status. Using an Inception-V1 network as feature extraction, our models achieve a Fréchet Inception Distance (FID) of 18.4. We find that using pathology meaningful features on these metrics show consistent performance, with a FID of 8.21. Furthermore, we asked two expert pathologists to distinguish our generated images from real ones, finding no significant difference between them.

Machine Learning algorithms are typically regarded as appropriate optimization schemes for minimizing risk functions that are constructed on the training set, which conveys statistical flavor to the corresponding learning problem. When the focus is shifted on perception, which is inherently interwound with time, recent alternative formulations of learning have been proposed that rely on the principle of Least Cognitive Action, which very much reminds us of the Least Action Principle in mechanics. In this paper, we discuss different forms of the cognitive action and show the well-posedness of learning. In particular, unlike the special case of the action in mechanics, where the stationarity is typically gained on saddle points, we prove the existence of the minimum of a special form of cognitive action, which yields forth-order differential equations of learning. We also briefly discuss the dissipative behavior of these equations that turns out to characterize the process of learning.

In linear inverse problems, the goal is to recover a target signal from undersampled, incomplete or noisy linear measurements. Typically, the recovery relies on complex numerical optimization methods; recent approaches perform an unfolding of a numerical algorithm into a neural network form, resulting in a substantial reduction of the computational complexity. In this paper, we consider the recovery of a target signal with the aid of a correlated signal, the so-called side information (SI), and propose a deep unfolding model that incorporates SI. The proposed model is used to learn coupled representations of correlated signals from different modalities, enabling the recovery of multimodal data at a low computational cost. As such, our work introduces the first deep unfolding method with SI, which actually comes from a different modality. We apply our model to reconstruct near-infrared images from undersampled measurements given RGB images as SI. Experimental results demonstrate the superior performance of the proposed framework against single-modal deep learning methods that do not use SI, multimodal deep learning designs, and optimization algorithms.

Statistical machine learning models should be evaluated and validated before putting to work. Conventional k-fold Monte Carlo Cross-Validation (MCCV) procedure uses a pseudo-random sequence to partition instances into k subsets, which usually causes subsampling bias, inflates generalization errors and jeopardizes the reliability and effectiveness of cross-validation. Based on ordered systematic sampling theory in statistics and low-discrepancy sequence theory in number theory, we propose a new k-fold cross-validation procedure by replacing a pseudo-random sequence with a best-discrepancy sequence, which ensures low subsampling bias and leads to more precise Expected-Prediction-Error estimates. Experiments with 156 benchmark datasets and three classifiers (logistic regression, decision tree and naive bayes) show that in general, our cross-validation procedure can extrude subsampling bias in the MCCV by lowering the EPE around 7.18% and the variances around 26.73%. In comparison, the stratified MCCV can reduce the EPE and variances of the MCCV around 1.58% and 11.85% respectively. The Leave-One-Out (LOO) can lower the EPE around 2.50% but its variances are much higher than the any other CV procedure. The computational time of our cross-validation procedure is just 8.64% of the MCCV, 8.67% of the stratified MCCV and 16.72% of the LOO. Experiments also show that our approach is more beneficial for datasets characterized by relatively small size and large aspect ratio. This makes our approach particularly pertinent when solving bioscience classification problems. Our proposed systematic subsampling technique could be generalized to other machine learning algorithms that involve random subsampling mechanism.

Canonical Correlation Analysis (CCA) is a classic technique for multi-view data analysis. To overcome the deficiency of linear correlation in practical multi-view learning tasks, various CCA variants were proposed to capture nonlinear dependency. However, it is non-trivial to have an in-principle understanding of these variants due to their inherent restrictive assumption on the data and latent code distributions. Although some works have studied probabilistic interpretation for CCA, these models still require the explicit form of the distributions to achieve a tractable solution for the inference. In this work, we study probabilistic interpretation for CCA based on implicit distributions. We present Conditional Mutual Information (CMI) as a new criterion for CCA to consider both linear and nonlinear dependency for arbitrarily distributed data. To eliminate direct estimation for CMI, in which explicit form of the distributions is still required, we derive an objective which can provide an estimation for CMI with efficient inference methods. To facilitate Bayesian inference of multi-view analysis, we propose Adversarial CCA (ACCA), which achieves consistent encoding for multi-view data with the consistent constraint imposed on the marginalization of the implicit posteriors. Such a model would achieve superiority in the alignment of the multi-view data with implicit distributions. It is interesting to note that most of the existing CCA variants can be connected with our proposed CCA model by assigning specific form for the posterior and likelihood distributions. Extensive experiments on nonlinear correlation analysis and cross-view generation on benchmark and real-world datasets demonstrate the superiority of our model.

In this paper, we introduce a novel method to generate interpretable regression function estimators. The idea is based on called data-dependent coverings. The aim is to extract from the data a covering of the feature space instead of a partition. The estimator predicts the empirical conditional expectation over the cells of the partitions generated from the coverings. Thus, such estimator has the same form as those issued from data-dependent partitioning algorithms. We give sufficient conditions to ensure the consistency, avoiding the sufficient condition of shrinkage of the cells that appears in the former literature. Doing so, we reduce the number of covering elements. We show that such coverings are interpretable and each element of the covering is tagged as significant or insignificant. The proof of the consistency is based on a control of the error of the empirical estimation of conditional expectations which is interesting on its own.

We address the problem of unsupervised domain adaptation (UDA) by learning a cross-domain agnostic embedding space, where the distance between the probability distributions of the two source and target visual domains is minimized. We use the output space of a shared cross-domain deep encoder to model the embedding space anduse the Sliced-Wasserstein Distance (SWD) to measure and minimize the distance between the embedded distributions of two source and target domains to enforce the embedding to be domain-agnostic.Additionally, we use the source domain labeled data to train a deep classifier from the embedding space to the label space to enforce the embedding space to be discriminative.As a result of this training scheme, we provide an effective solution to train the deep classification network on the source domain such that it will generalize well on the target domain, where only unlabeled training data is accessible. To mitigate the challenge of class matching, we also align corresponding classes in the embedding space by using high confidence pseudo-labels for the target domain, i.e. assigning the class for which the source classifier has a high prediction probability. We provide theoretical justification as well as experimental results on UDA benchmark tasks to demonstrate that our method is effective and leads to state-of-the-art performance.

Research on style transfer and domain translation has clearly demonstrated the ability of deep learning-based algorithms to manipulate images in terms of artistic style. More recently, several attempts have been made to extend such approaches to music (both symbolic and audio) in order to enable transforming musical style in a similar manner. In this study, we focus on symbolic music with the goal of altering the 'style' of a piece while keeping its original 'content'. As opposed to the current methods, which are inherently restricted to be unsupervised due to the lack of 'aligned' data (i.e. the same musical piece played in multiple styles), we develop the first fully supervised algorithm for this task. At the core of our approach lies a synthetic data generation scheme which allows us to produce virtually unlimited amounts of aligned data, and hence avoid the above issue. In view of this data generation scheme, we propose an encoder-decoder model for translating symbolic music accompaniments between a number of different styles. Our experiments show that our models, although trained entirely on synthetic data, are capable of producing musically meaningful accompaniments even for real (non-synthetic) MIDI recordings.

Connections between integration along hypersufaces, Radon transforms, and neural networks are exploited to highlight an integral geometric mathematical interpretation of neural networks. By analyzing the properties of neural networks as operators on probability distributions for observed data, we show that the distribution of outputs for any node in a neural network can be interpreted as a nonlinear projection along hypersurfaces defined by level surfaces over the input data space. We utilize these descriptions to provide new interpretation for phenomena such as nonlinearity, pooling, activation functions, and adversarial examples in neural network-based learning problems.

Machine learning is often used in virtual screening to find compounds that are pharmacologically active on a target protein. The weave module is a type of graph convolutional deep neural network that uses not only features focusing on atoms alone (atom features) but also features focusing on atom pairs (pair features); thus, it can consider information of nonadjacent atoms. However, the correlation between the distance on the graph and the three-dimensional coordinate distance is uncertain. In this paper, we propose three improvements for modifying the weave module. First, the distances between ring atoms on the graph were modified to bring the distances on the graph closer to the coordinate distance. Second, different weight matrices were used depending on the distance on the graph in the convolution layers of the pair features. Finally, a weighted sum, by distance, was used when converting pair features to atom features. The experimental results show that the performance of the proposed method is slightly better than that of the weave module, and the improvement in the distance representation might be useful for compound activity prediction.