Selecting ad creative is one of the most important task for DSPs (Demand-Side Platform) in online advertising. DSPs should not only consider the effectiveness of the ad creative but also the user's psychological status when selecting ad creative. In this study, we propose an efficient and easy-to-implement ad creative selection algorithm that explicitly considers wear-in and wear-out effects of ad creative due to the repetitive ad exposures. The proposed system was deployed in a real-world production environment and tested against the baseline. It out-performed the existing system in most of the KPIs.

In-bed pose estimation has shown value in fields such as hospital patient monitoring, sleep studies, and smart homes. In this paper, we present a novel in-bed pressure-based pose estimation approach capable of accurately detecting body parts from highly ambiguous pressure data. We exploit the idea of using a learnable pre-processing step, which transforms the vague pressure maps to a representation close to the expected input space of common purpose pose identification modules, which fail if solely used on the pressure data. To this end, a fully convolutional network with multiple scales is used as the learnable pre-processing step to provide the pose-specific characteristics of the pressure maps to the pre-trained pose identification module. A combination of loss functions is used to model the constraints, ensuring that unclear body parts are reconstructed correctly while preventing the pre-processing block from generating arbitrary images. The evaluation results show high visual fidelity in the generated pre-processed images as well as high detection rates in pose estimation. Furthermore, we show that the trained pre-processing block can be effective for pose identification models for which it has not been trained as well.

Improving a semi-supervised image segmentation task has the option of adding more unlabelled images, labelling the unlabelled images or combining both, as neither image acquisition nor expert labelling can be considered trivial in most clinical applications. With a laparoscopic liver image segmentation application, we investigate the performance impact by altering the quantities of labelled and unlabelled training data, using a semi-supervised segmentation algorithm based on the mean teacher learning paradigm. We first report a significantly higher segmentation accuracy, compared with supervised learning. Interestingly, this comparison reveals that the training strategy adopted in the semi-supervised algorithm is also responsible for this observed improvement, in addition to the added unlabelled data. We then compare different combinations of labelled and unlabelled data set sizes for training semi-supervised segmentation networks, to provide a quantitative example of the practically useful trade-off between the two data planning strategies in this surgical guidance application.

--Deep neural networks (DNNs) have demonstrated impressive performance on many challenging machine learning tasks. However, DNNs are vulnerable to adversarial inputs generated by adding maliciously crafted perturbations to the benign inputs. As a growing number of attacks have been reported to generate adversarial inputs of varying sophistication, the defense-attack arms race has been accelerated. MODEF intelligently combines unsupervised model denoising ensemble with supervised model verification ensemble by quantifying model diversity, aiming to boost the robustness of the target model against adversarial examples. Evaluated using eleven representative attacks on popular benchmark datasets, we show that MODEF achieves remarkable defense success rates, compared with existing defense methods, and provides a superior capability of repairing adversarial inputs and making correct predictions with high accuracy in the presence of black-box attacks. The recent advances in deep neural networks (DNNs) have powered numerous applications in different domains due to their outstanding performance compared to traditional machine learning techniques. However, it has been shown that DNNs can be easily fooled by adversarial inputs [1], making them become a double-edged sword as the vulnerability of DNNs to adversarial attacks has posed serious threats to many security-critical applications, such as biometric authentication and autonomous driving. As a number of defenses are being proposed, more attacks of varying sophistication have been put forward, accelerating the defense-attack arms race. Some even argue that designing new attacks requires much less efforts than developing effective defenses. Thus, improving the robustness and defensibility against adversarial attacks is crucial. Adversarial examples are generated by maliciously perturbing benign examples sent to the target DNN model through querying its prediction API, aiming to fool and mislead the target model to misclassify by producing incorrect predictions randomly (untargeted attack) or purposefully (targeted attack).

The developments of Rademacher complexity and PAC-Bayesian theory have been largely independent. One exception is the PAC-Bayes theorem of Kakade, Sridharan, and Tewari (2008), which is established via Rademacher complexity theory by viewing Gibbs classifiers as linear operators. The goal of this paper is to extend this bridge between Rademacher complexity and state-of-the-art PAC-Bayesian theory. We first demonstrate that one can match the fast rate of Catoni's PAC-Bayes bounds (Catoni, 2007) using shifted Rademacher processes (Wegkamp, 2003; Lecu\'{e} and Mitchell, 2012; Zhivotovskiy and Hanneke, 2018). We then derive a new fast-rate PAC-Bayes bound in terms of the "flatness" of the empirical risk surface on which the posterior concentrates. Our analysis establishes a new framework for deriving fast-rate PAC-Bayes bounds and yields new insights on PAC-Bayesian theory.

Graph neural networks (GNNs) are widely used in many applications. However, their robustness against adversarial attacks is criticized. Prior studies show that using unnoticeable modifications on graph topology or nodal features can significantly reduce the performances of GNNs. It is very challenging to design robust graph neural networks against poisoning attack and several efforts have been taken. Existing work aims at reducing the negative impact from adversarial edges only with the poisoned graph, which is sub-optimal since they fail to discriminate adversarial edges from normal ones. On the other hand, clean graphs from similar domains as the target poisoned graph are usually available in the real world. By perturbing these clean graphs, we create supervised knowledge to train the ability to detect adversarial edges so that the robustness of GNNs is elevated. However, such potential for clean graphs is neglected by existing work. To this end, we investigate a novel problem of improving the robustness of GNNs against poisoning attacks by exploring clean graphs. Specifically, we propose PA-GNN, which relies on a penalized aggregation mechanism that directly restrict the negative impact of adversarial edges by assigning them lower attention coefficients. To optimize PA-GNN for a poisoned graph, we design a meta-optimization algorithm that trains PA-GNN to penalize perturbations using clean graphs and their adversarial counterparts, and transfers such ability to improve the robustness of PA-GNN on the poisoned graph. Experimental results on four real-world datasets demonstrate the robustness of PA-GNN against poisoning attacks on graphs.

Pre-season prediction of crop production outcomes such as grain yields and N losses can provide insights to stakeholders when making decisions. Simulation models can assist in scenario planning, but their use is limited because of data requirements and long run times. Thus, there is a need for more computationally expedient approaches to scale up predictions. We evaluated the potential of five machine learning (ML) algorithms as meta-models for a cropping systems simulator (APSIM) to inform future decision-support tool development. We asked: 1) How well do ML meta-models predict maize yield and N losses using pre-season information? 2) How many data are needed to train ML algorithms to achieve acceptable predictions?; 3) Which input data variables are most important for accurate prediction?; and 4) Do ensembles of ML meta-models improve prediction? The simulated dataset included more than 3 million genotype, environment and management scenarios. Random forests most accurately predicted maize yield and N loss at planting time, with a RRMSE of 14% and 55%, respectively. ML meta-models reasonably reproduced simulated maize yields but not N loss. They also differed in their sensitivities to the size of the training dataset. Across all ML models, yield prediction error decreased by 10-40% as the training dataset increased from 0.5 to 1.8 million data points, whereas N loss prediction error showed no consistent pattern. ML models also differed in their sensitivities to input variables. Averaged across all ML models, weather conditions, soil properties, management information and initial conditions were roughly equally important when predicting yields. Modest prediction improvements resulted from ML ensembles. These results can help accelerate progress in coupling simulation models and ML toward developing dynamic decision support tools for pre-season management.

Aggregating multiple learners through an ensemble of models aims to make better predictions by capturing the underlying distribution more accurately. Different ensembling methods, such as bagging, boosting and stacking/blending, have been studied and adopted extensively in research and practice. While bagging and boosting intend to reduce variance and bias, respectively, blending approaches target both by finding the optimal way to combine base learners to find the best trade-off between bias and variance. In blending, ensembles are created from weighted averages of multiple base learners. In this study, a systematic approach is proposed to find the optimal weights to create these ensembles for bias-variance tradeoff using cross-validation for regression problems (Cross-validated Optimal Weighted Ensemble (COWE)). Furthermore, it is known that tuning hyperparameters of each base learner inside the ensemble weight optimization process can produce better performing ensembles. To this end, a nested algorithm based on bi-level optimization that considers tuning hyperparameters as well as finding the optimal weights to combine ensembles (Cross-validated Optimal Weighted Ensemble with Internally Tuned Hyperparameters (COWE-ITH)) was proposed. The algorithm is shown to be generalizable to real data sets though analyses with ten publicly available data sets. The prediction accuracies of COWE-ITH and COWE have been compared to base learners and the state-of-art ensemble methods. The results show that COWE-ITH outperforms other benchmarks as well as base learners in 9 out of 10 data sets.

Circadian rhythms govern most essential biological processes in the human body; they influence multiple biological activities including sleep, performance, mood, skin temperature, hormone production, and nutrient absorption. The dim light melatonin onset (DLMO) is the current gold standard for measuring human circadian phase (or timing). The collection of DLMO is expensive and time consuming since multiple saliva or blood samples are required from nighttime studies in specialized environmental conditions. In the past few years, several non-invasive approaches have been designed for estimating DLMO values. These methods collect daily sampled data (e.g., sleep onset/offset times) or frequently sampled data (e.g., light exposure, skin temperature, physical activity collected every minute) to train learning models for estimating DLMO, therefore previous studies only leveraged one time scale. In this paper, we propose a two-step framework for estimating DLMO using the data of both time scales. The first step summarizes the data prior to the current day, while the second step combines this summary with frequently sampled data of the current day. We evaluate several variants of moving average model which input sleep timing data as the first step and recurrent neural network models as the second step for estimating DLMO. The experimental results show that our two-step model with two-time-scale features has statistically significantly lower root-mean-square errors than the models that use either daily sampled data or frequently sampled data alone.

We propose a novel high-performance interpretable deep tabular data learning network, TabNet. TabNet utilizes a sequential attention mechanism to choose which features to reason from at each decision step and then aggregates the processed information towards the final decision. Explicit selection of sparse features enables more efficient learning as the model capacity at each decision step is fully utilized for the most relevant features, and also more interpretable decision making via visualization of selection masks. We demonstrate that TabNet outperforms other neural network and decision tree variants on a wide range of tabular data learning datasets while yielding interpretable feature attributions and insights into the global model behavior.