Recently, Qiao, Duan, and Cheng~(2019) proposed a distributed nearest-neighbor classification method, in which a massive dataset is split into smaller groups, each processed with a $k$-nearest-neighbor classifier, and the final class label is predicted by a majority vote among these groupwise class labels. This paper shows that the distributed algorithm with $k=1$ over a sufficiently large number of groups attains a minimax optimal error rate up to a multiplicative logarithmic factor under some regularity conditions, for both regression and classification problems. Roughly speaking, distributed 1-nearest-neighbor rules with $M$ groups has a performance comparable to standard $\Theta(M)$-nearest-neighbor rules. In the analysis, alternative rules with a refined aggregation method are proposed and shown to attain exact minimax optimal rates.

Optical imaging is commonly used for both scientific and technological applications across industry and academia. In image sensing, a measurement, such as of an object's position, is performed by computational analysis of a digitized image. An emerging image-sensing paradigm breaks this delineation between data collection and analysis by designing optical components to perform not imaging, but encoding. By optically encoding images into a compressed, low-dimensional latent space suitable for efficient post-analysis, these image sensors can operate with fewer pixels and fewer photons, allowing higher-throughput, lower-latency operation. Optical neural networks (ONNs) offer a platform for processing data in the analog, optical domain. ONN-based sensors have however been limited to linear processing, but nonlinearity is a prerequisite for depth, and multilayer NNs significantly outperform shallow NNs on many tasks. Here, we realize a multilayer ONN pre-processor for image sensing, using a commercial image intensifier as a parallel optoelectronic, optical-to-optical nonlinear activation function. We demonstrate that the nonlinear ONN pre-processor can achieve compression ratios of up to 800:1 while still enabling high accuracy across several representative computer-vision tasks, including machine-vision benchmarks, flow-cytometry image classification, and identification of objects in real scenes. In all cases we find that the ONN's nonlinearity and depth allowed it to outperform a purely linear ONN encoder. Although our experiments are specialized to ONN sensors for incoherent-light images, alternative ONN platforms should facilitate a range of ONN sensors. These ONN sensors may surpass conventional sensors by pre-processing optical information in spatial, temporal, and/or spectral dimensions, potentially with coherent and quantum qualities, all natively in the optical domain.

Missing values, widely called as \textit{sparsity} in literature, is a common characteristic of many real-world datasets. Many imputation methods have been proposed to address this problem of data incompleteness or sparsity. However, the accuracy of a data imputation method for a given feature or a set of features in a dataset is highly dependent on the distribution of the feature values and its correlation with other features. Another problem that plagues industry deployments of machine learning (ML) solutions is concept drift detection, which becomes more challenging in the presence of missing values. Although data imputation and concept drift detection have been studied extensively, little work has attempted a combined study of the two phenomena, i.e., concept drift detection in the presence of sparsity. In this work, we carry out a systematic study of the following: (i) different patterns of missing values, (ii) various statistical and ML based data imputation methods for different kinds of sparsity, (iii) several concept drift detection methods, (iv) practical analysis of the various drift detection metrics, (v) selecting the best concept drift detector given a dataset with missing values based on the different metrics. We first analyze it on synthetic data and publicly available datasets, and finally extend the findings to our deployed solution of automated change risk assessment system. One of the major findings from our empirical study is the absence of supremacy of any one concept drift detection method across all the relevant metrics. Therefore, we adopt a majority voting based ensemble of concept drift detectors for abrupt and gradual concept drifts. Our experiments show optimal or near optimal performance can be achieved for this ensemble method across all the metrics.

Hyperparameter optimization constitutes a large part of typical modern machine learning workflows. This arises from the fact that machine learning methods and corresponding preprocessing steps often only yield optimal performance when hyperparameters are properly tuned. But in many applications, we are not only interested in optimizing ML pipelines solely for predictive accuracy; additional metrics or constraints must be considered when determining an optimal configuration, resulting in a multi-objective optimization problem. This is often neglected in practice, due to a lack of knowledge and readily available software implementations for multi-objective hyperparameter optimization. In this work, we introduce the reader to the basics of multi-objective hyperparameter optimization and motivate its usefulness in applied ML. Furthermore, we provide an extensive survey of existing optimization strategies, both from the domain of evolutionary algorithms and Bayesian optimization. We illustrate the utility of MOO in several specific ML applications, considering objectives such as operating conditions, prediction time, sparseness, fairness, interpretability and robustness.

Analyzing classification model performance is a crucial task for machine learning practitioners. While practitioners often use count-based metrics derived from confusion matrices, like accuracy, many applications, such as weather prediction, sports betting, or patient risk prediction, rely on a classifier's predicted probabilities rather than predicted labels. In these instances, practitioners are concerned with producing a calibrated model, that is, one which outputs probabilities that reflect those of the true distribution. Model calibration is often analyzed visually, through static reliability diagrams, however, the traditional calibration visualization may suffer from a variety of drawbacks due to the strong aggregations it necessitates. Furthermore, count-based approaches are unable to sufficiently analyze model calibration. We present Calibrate, an interactive reliability diagram that addresses the aforementioned issues. Calibrate constructs a reliability diagram that is resistant to drawbacks in traditional approaches, and allows for interactive subgroup analysis and instance-level inspection. We demonstrate the utility of Calibrate through use cases on both real-world and synthetic data. We further validate Calibrate by presenting the results of a think-aloud experiment with data scientists who routinely analyze model calibration.

In many object recognition applications, the set of possible categories is an open set, and the deployed recognition system will encounter novel objects belonging to categories unseen during training. Detecting such "novel category" objects is usually formulated as an anomaly detection problem. Anomaly detection algorithms for feature-vector data identify anomalies as outliers, but outlier detection has not worked well in deep learning. Instead, methods based on the computed logits of visual object classifiers give state-of-the-art performance. This paper proposes the Familiarity Hypothesis that these methods succeed because they are detecting the absence of familiar learned features rather than the presence of novelty. This distinction is important, because familiarity-based detection will fail in many situations where novelty is present. For example when an image contains both a novel object and a familiar one, the familiarity score will be high, so the novel object will not be noticed. The paper reviews evidence from the literature and presents additional evidence from our own experiments that provide strong support for this hypothesis. The paper concludes with a discussion of whether familiarity-based detection is an inevitable consequence of representation learning.

Graph Neural Networks (GNNs), inspired by Convolutional Neural Networks (CNNs), aggregate the message of nodes' neighbors and structure information to acquire expressive representations of nodes for node classification, graph classification, and link prediction. Previous studies have indicated that GNNs are vulnerable to Membership Inference Attacks (MIAs), which infer whether a node is in the training data of GNNs and leak the node's private information, like the patient's disease history. The implementation of previous MIAs takes advantage of the models' probability output, which is infeasible if GNNs only provide the prediction label (label-only) for the input. In this paper, we propose a label-only MIA against GNNs for node classification with the help of GNNs' flexible prediction mechanism, e.g., obtaining the prediction label of one node even when neighbors' information is unavailable. Our attacking method achieves around 60\% accuracy, precision, and Area Under the Curve (AUC) for most datasets and GNN models, some of which are competitive or even better than state-of-the-art probability-based MIAs implemented under our environment and settings. Additionally, we analyze the influence of the sampling method, model selection approach, and overfitting level on the attack performance of our label-only MIA. Both of those factors have an impact on the attack performance. Then, we consider scenarios where assumptions about the adversary's additional dataset (shadow dataset) and extra information about the target model are relaxed. Even in those scenarios, our label-only MIA achieves a better attack performance in most cases. Finally, we explore the effectiveness of possible defenses, including Dropout, Regularization, Normalization, and Jumping knowledge. None of those four defenses prevent our attack completely.

We present a novel machine learning based approach for detecting galaxy-scale gravitational lenses from interferometric data, specifically those taken with the International LOFAR Telescope (ILT), which is observing the northern radio sky at a frequency of 150 MHz, an angular resolution of 350 mas and a sensitivity of 90 µJy beam 1 (1σ). We develop and test several Convolutional Neural Networks to determine the probability and uncertainty of a given sample being classified as a lensed or non-lensed event. By training and testing on a simulated interferometric imaging data set that includes realistic lensed and non-lensed radio sources, we find that it is possible to recover 95.3 per cent of the lensed samples (true positive rate), with a contamination of just 0.008 per cent from non-lensed samples (false positive rate). Taking the expected lensing probability into account results in a predicted sample purity for lensed events of 92.2 per cent. We find that the network structure is most robust when the maximum image separation between the lensed images is 3 times the synthesized beam size, and the lensed images have a total flux density that is equivalent to at least a 20σ (point-source) detection.

To evaluate the use of artificial intelligence (AI) to shorten digital breast tomosynthesis (DBT) reading time while maintaining or improving accuracy. A deep learning AI system was developed to identify suspicious soft-tissue and calcified lesions in DBT images. A reader study compared the performance of 24 radiologists (13 of whom were breast subspecialists) reading 260 DBT examinations (including 65 cancer cases) both with and without AI. Readings occurred in two sessions separated by at least 4 weeks. Area under the receiver operating characteristic curve (AUC), reading time, sensitivity, specificity, and recall rate were evaluated with statistical methods for multireader, multicase studies. Radiologist performance for the detection of malignant lesions, measured by mean AUC, increased 0.057 with the use of AI (95% confidence interval [CI]: 0.028, 0.087; P .01), Reading time decreased 52.7% (95% CI: 41.8%, 61.5%; P .01), Sensitivity increased from 77.0% without AI to 85.0% with AI (8.0%; 95% CI: 2.6%, 13.4%; P .01), The concurrent use of an accurate DBT AI system was found to improve cancer detection efficacy in a reader study that demonstrated increases in AUC, sensitivity, and specificity and a reduction in recall rate and reading time. See also the commentary by Hsu and Hoyt in this issue. Reading times were significantly reduced, and sensitivity, specificity, and recall rate improved in a nonclinical reader study when an artificial intelligence system was utilized concurrently with image interpretation for digital breast tomosynthesis.

How do international crises unfold? We conceptualize of international relations as a strategic chess game between adversaries and develop a systematic way to measure pieces, moves, and gambits accurately and consistently over a hundred years of history. We introduce a new ontology and dataset of international events called ICBe based on a very high-quality corpus of narratives from the International Crisis Behavior (ICB) Project. We demonstrate that ICBe has higher coverage, recall, and precision than existing state of the art datasets and conduct two detailed case studies of the Cuban Missile Crisis (1962) and Crimea-Donbas Crisis (2014). We further introduce two new event visualizations (event icongraphy and crisis maps), an automated benchmark for measuring event recall using natural language processing (sythnetic narratives), and an ontology reconstruction task for objectively measuring event precision. We make the data, online appendix, replication material, and visualizations of every historical episode available at a companion website www.crisisevents.org and the github repository.