Out-of-Distribution (OOD) extrapolation, where test data feature values extend beyond the training range, poses significant challenges in machine learning. While existing solutions often sacrifice interpretability, resulting in limited applicability in high-stakes applications where interpretability is a a critical requirement. In this paper, we propose Delta-GAM, an interpretable Generalized Additive Model (GAM) that achieves robust extrapolation in OOD scenarios. Our method jointly learns (1) feature-target relationships and (2) functional adaptations for extrapolating beyond the training distribution by reformulating GAM fitting as a second-order interaction problem between features and their distributional offsets. We theoretically show that smooth GAM shape functions induce an approximately low-rank structure in these interactions, enabling efficient decomposition via a specialized neural network. Experiments on synthetic and real-world data demonstrate Delta-GAM's superior performance in OOD extrapolation tasks while preserving model interpretability, bridging a key gap in trustworthy machine learning.

In this paper, we consider a typical microgrid scenario that consists of centralized power grid, renewable energy generation, and combined heat and power (CHP) local (co-)generation, as well as power and heat energy storage devices. We aim to minimize the microgrid's operating cost by formulating it as a stochastic non-convex optimization programming, which is challenging to solve optimally. We design an online algorithm by developing a modified Lyapunov optimization approach based on the random system inputs (e.g., the acquired electricity from power grid, the charging/discharging of the energy storage devices, obtained power from the local generator, and the renewable energy generation etc.), which does not require any statistic information of the system. Considering that the nonconvexity of the problem is caused by the dependence of power in battery pack and heat energy in thermal tank, we further explore the relation between them and convert the problem into a convex stochastic optimization programming. We show that the proposed algorithm is efficient with very low computational complexity and is proved to achieve near optimal performance. Moreover, extensive empirical evaluations using real-world traces are provided to study the effectiveness of the proposed algorithm.

Social network has become a very popular way for internet users to communicate and interact online. Users spend plenty of time on famous social networks (e.g., Facebook, Twitter, Sina Weibo, etc.), reading news, discussing events and posting messages. Unfortunately, this popularity also attracts a significant amount of spammers who continuously expose malicious behavior (e.g., post messages containing commercial URLs, following a larger amount of users, etc.), leading to great misunderstanding and inconvenience on users׳ social activities. In this paper, a supervised machine learning based solution is proposed for an effective spammer detection. The main procedure of the work is: first, collect a dataset from Sina Weibo including 30,116 users and more than 16 million messages. Then, construct a labeled dataset of users and manually classify users into spammers and non-spammers. Afterwards, extract a set of feature from message content and users׳ social behavior, and apply into SVM (Support Vector Machines) based spammer detection algorithm. The experiment shows that the proposed solution is capable to provide excellent performance with true positive rate of spammers and non-spammers reaching 99.1% and 99.9% respectively.

Social network has become a very popular way for internet users to communicate and interact online. Users spend a great deal of time on famous social networks (e.g. Facebook, Twitter, Sina Weibo, etc.), reading news, discussing events and posting their messages. Unfortunately, this popularity also attracts a significant amount of spammers who continuously expose malicious behaviors (e.g. Post messages containing commercial topics or URLs, following a larger amount of users, etc.), leading to great inconvenience on normal users' social activities. In this paper, a supervised machine learning based spammer filtering method is proposed. We first collected a dataset from Sina Weibo that includes 30,116 users and more than 16 million messages, then, construct a labeled dataset of users and manually classify users into spammers and non-spammers, after that, abstract a set of novel features from message content and users' social behavior, and apply into SVM based spammer classifier. Our experiments show that true positive rate of spammers and non-spammers could reach 99.1% and 99.9%.

Design Space Exploration (DSE) is one of the most important stages in High Level Synthesis designing methodology. This paper presents a novel DSE approach for the current generation of systems with heterogeneous multi parametric optimization objectives. The method introduced in this paper is capable of concurrently resolving multiple conflicting issues encountered during DSE, such as maximization of accuracy needed in the evaluation of design space with minimization in time expended to explore the best architecture. Results of the proposed method for different benchmarks indicated significant acceleration in exploration process compared to another existing approach that is also based on Pareto optimal analysis.

In High level Synthesis design methodology, the evaluation and selection of the optimal architecture for the system is done through a process called Design Space Exploration (DSE). This paper presents a novel framework for fast DSE using fuzzy search technique for optimizing modular computing architecture for the current generation of multi objective VLSI designs. The proposed method is able to radically reduce the number of architectural variants to be analyzed during design space exploration while simultaneously maintaining the precision required during the exploration process. Significant improvement in speedup during DSE is obtained for different benchmarks, compared to a DSE method with binary search mechanism.