Hyperparameters are critical in machine learning, as different hyperparameters often result in models with significantly different performance. Hyperparameters may be deemed confidential because of their commercial value and the confidentiality of the proprietary algorithms that the learner uses to learn them. In this work, we propose attacks on stealing the hyperparameters that are learnt by a learner. We call our attacks hyperparameter stealing attacks. Our attacks are applicable to a variety of popular machine learning algorithms such as ridge regression, logistic regression, support vector machine, and neural network. We evaluate the effectiveness of our attacks both theoretically and empirically. For instance, we evaluate our attacks on Amazon Machine Learning. Our results demonstrate that our attacks can accurately steal hyperparameters. We also study countermeasures. Our results highlight the need for new defenses against our hyperparameter stealing attacks for certain machine learning algorithms.

Background-␤-Adrenergic signaling is downregulated in the failing heart, and the significance of such change remains unclear. Methods and Results-To address the role of ␤-adrenergic dysfunction in heart failure (HF), aortic stenosis (AS) was induced in wild-type (WT) and transgenic (TG) mice with cardiac targeted overexpression of ␤ 2 -adrenergic receptors (ARs), and animals were studied 9 weeks later. The extents of increase in systolic arterial pressure (PϽ0.01 versus controls), left ventricular (LV) hypertrophy (TG, 94Ϯ6 to 175Ϯ7 mg; WT, 110Ϯ6 to 168Ϯ10 mg; both PϽ0.01), and expression of ANP mRNA were similar between TG and WT mice with AS. TG mice had higher incidences of premature death and critical illness due to heart failure (75% versus 23%), pleural effusion (81% versus 45%), and left atrial thrombosis (81% versus 36%, all PϽ0.05). A more extensive focal fibrosis was found in the hypertrophied LV of TG mice (PϽ0.05). These findings indicate a more severe LV dysfunction in TG mice. In sham-operated mice, LV dP/dt max and heart rate were markedly higher in TG than WT mice (both PϽ0.01). dP/dt max was lower in both AS groups than in sham-operated controls, and this tended to be more pronounced in TG than WT mice (Ϫ32Ϯ5% versus Ϫ16Ϯ6%, Pϭ0.059), although dP/dt max remained higher in TG than WT groups (PϽ0.05). Conclusions-Elevated

Community detection plays a key role in understanding graph structure. However, several recent studies showed that community detection is vulnerable to adversarial structural perturbation. In particular, via adding or removing a small number of carefully selected edges in a graph, an attacker can manipulate the detected communities. However, to the best of our knowledge, there are no studies on certifying robustness of community detection against such adversarial structural perturbation. In this work, we aim to bridge this gap. Specifically, we develop the first certified robustness guarantee of community detection against adversarial structural perturbation. Given an arbitrary community detection method, we build a new smoothed community detection method via randomly perturbing the graph structure. We theoretically show that the smoothed community detection method provably groups a given arbitrary set of nodes into the same community (or different communities) when the number of edges added/removed by an attacker is bounded. Moreover, we show that our certified robustness is tight. We also empirically evaluate our method on multiple real-world graphs with ground truth communities.

No abstract

A high level of beta(2)AR overexpression results in cardiomyopathy and heart failure. The onset was slower and the expression levels of receptors required are much higher than previously described for the beta(1)AR overexpression.