Hui Chen

The changesets (or patches) that fix open source software vulnerabilities form critical datasets for various machine learning security-enhancing applications, such as automated vulnerability patching and silent fix detection. These patch datasets are derived from extensive collections of historical vulnerability fixes, maintained in databases like the Common Vulnerabilities and Exposures list and the National Vulnerability Database. However, since these databases focus on rapid notification to the security community, they contain significant inaccuracies and omissions that have a negative impact on downstream software security quality assurance tasks. In this paper, we propose an approach employing Uncertainty Quantification (UQ) to curate datasets of publicly-available software vulnerability patches. Our methodology leverages machine learning models that incorporate UQ to differentiate between patches based on their potential utility. We begin by evaluating a number of popular UQ techniques, including Vanilla, Monte Carlo Dropout, and Model Ensemble, as well as homoscedastic and heteroscedastic models of noise. Our findings indicate that Model Ensemble and heteroscedastic models are the best choices for vulnerability patch datasets. Based on these UQ modeling choices, we propose a heuristic that uses UQ to filter out lower quality instances and select instances with high utility value from the vulnerability dataset. Using our approach, we observe an improvement in predictive performance and a significant reduction of model training time (i.e., energy consumption) for a state-of-the-art vulnerability prediction model.

This work stems from three observations on prior Just-In-Time Software Defect Prediction (JIT-SDP) models. First, prior studies treat the JIT-SDP problem solely as a classification problem. Second, prior JIT-SDP studies do not consider that class balancing processing may change the underlying characteristics of software changeset data. Third, only a single source of concept drift, the class imbalance evolution is addressed in prior JIT-SDP incremental learning models. We propose an incremental learning framework called CPI-JIT for JIT-SDP. First, in addition to a classification modeling component, the framework includes a time-series forecast modeling component in order to learn temporal interdependent relationship in the changesets. Second, the framework features a purposefully designed over-sampling balancing technique based on SMOTE and Principal Curves called SMOTE-PC. SMOTE-PC preserves the underlying distribution of software changeset data. In this framework, we propose an incremental deep neural network model called DeepICP. Via an evaluation using \numprojs software projects, we show that: 1) SMOTE-PC improves the model's predictive performance; 2) to some software projects it can be beneficial for defect prediction to harness temporal interdependent relationship of software changesets; and 3) principal curves summarize the underlying distribution of changeset data and reveals a new source of concept drift that the DeepICP model is proposed to adapt to.

In view of the opportunities and challenges brought by vocational education at the undergraduate level to professional teachers in higher vocational schools, combined with the current situation of teacher development in our school, we will explore the development path of professional teachers from the perspective of front-line teachers. Firstly, it analyses the hierarchical structure of professional teacher development, and analyses the stages that are easier to break through in the hierarchical structure. Then, it establishes the triggering system for professional teacher development, and demonstrates the necessity of triggering the internal driving force of teacher development from the perspective of classroom, emotion and environment. Finally, a professional teacher development accumulation matrix is established by taking engineering measurement technology professional teachers as an example, in order to learn from and inspire other vocational teachers in the period of rapid development, and keep in mind the original intention and mission together.

Soils are a complex ecosystem that provides critical services, such as growing food, supplying antibiotics, filtering wastes, and maintaining biodiversity; hence monitoring soil health and domestication is required for sustainable human development. Low-cost and high-resolution soil monitoring systems are challenging to design and build. Compounded by the sheer size of the monitoring area of interest and the variety of biological, chemical, and physical parameters to monitor, naive approaches to adding or scheduling more sensors will suffer from cost and scalability problems. We investigate a multi-robot sensing system integrated with an active learning-based predictive modeling technique. Taking advantage of advances in machine learning, the predictive model allows us to interpolate and predict soil attributes of interest from the data collected by sensors and soil surveys. The system provides high-resolution prediction when the modeling output is calibrated with static land-based sensors. The active learning modeling technique allows our system to be adaptive in data collection strategy for time-varying data fields, utilizing aerial and land robots for new sensor data. We evaluated our approach using numerical experiments with a soil dataset focusing on heavy metal concentration in a flooded area. The experimental results demonstrate that our algorithms can reduce sensor deployment costs via optimized sensing locations and paths while providing high-fidelity data prediction and interpolation. More importantly, the results verify the adapting behavior of the system to the spatial and temporal variations of soil conditions.

Recent years have experienced sustained focus in research on software defect prediction that aims to predict the likelihood of software defects. Moreover, with the increased interest in continuous deployment, a variant of software defect prediction called Just-in-Time Software Defect Prediction (JIT-SDP) focuses on predicting whether each incremental software change is defective. JIT-SDP is unique in that it consists of two interconnected data streams, one consisting of the arrivals of software changes stemming from design and implementation, and the other the (defective or clean) labels of software changes resulting from quality assurance processes. We present a systematic survey of 67 JIT-SDP studies with the objective to help researchers advance the state of the art in JIT-SDP and to help practitioners become familiar with recent progress. We summarize best practices in each phase of the JIT-SDP workflow, carry out a meta-analysis of prior studies, and suggest future research directions. Our meta-analysis of JIT-SDP studies indicates, among other findings, that the predictive performance correlates with change defect ratio, suggesting that JIT-SDP is most performant in projects that experience relatively high defect ratios. Future research directions for JIT-SDP include situating each technique into its application domain, reliability-aware JIT-SDP, and user-centered JIT-SDP.

Yanshanian magmatisms are intensive in the eastern Jiangnan orogen and can be divided into early (150–136 Ma) and late (136–120 Ma) stages. About these granitic rocks, the petrogenesis and associated tectonic setting are still hotly debated. The Taiping–Huangshan complex, a representative granitic rock in this region, is selected to perform geochemical studies. Three lithologic units have been identified in the Taiping–Huangshan complex: granodiorite, monzogranite, and alkali‐feldspar granite with zircon U–Pb dating results of 149–144 Ma, 148 Ma, and 135–131 Ma, respectively, belonging to the early stage for the granodiorite and monzogranite, and late stage for the alkali‐feldspar granite. The granodiorite and monzogranite have similar major and trace element characteristics with intermediate‐acidic SiO₂ contents ranging from 65.9 to 69.7 wt.%, low Mg# values of 35.8–38.4, enriched LREE, Pb, Th, and U contents, and weak Eu anomalies (δ_Eu = 0.71–0.79). The alkali‐feldspar granite is characterized by high SiO₂ contents (73.3–76.6 wt.%), low Mg# (2.6–10.0), enriched Th, U, Ta, Pb, and Nd, depleted high‐field‐strength elements, and pronounced negative Eu anomalies (δ_Eu = 0.01–0.19). The granodiorite and monzogranite have consistent ε_Hf(t) (−5.38 to −9.41) and δ¹⁸O (7.18‰ to 11.00‰) values, while the alkali‐feldspar granite has elevated ε_Hf(t) values (−0.96 to −6.48) and δ¹⁸O values (6.27‰ to 8.34‰). Integrated geochemical characteristics and batch partial melting modelling constrain that the early stage granitic rocks derived from ca. 8% batch melting of the Neoproterozoic juvenile crust, and the late‐stage granites derived from a similar crust source with a little contribution from mantle. From early to late stage, the tectonic setting translated from post‐orogenic to anorogenic and the later corresponded to a back‐arc extensional setting as increasing of the slab subducted angle of the Palaeo‐Pacific Plate.