Background: To explore the relationship between the pathological changes of the colon, terminal ileum, lung, liver and kidney, and the changes of Bax, PCNA and PAF in a rat model of NEC. Methods: One hundred and forty neonatal SD rats were randomly divided into NEC group and control group (70 in each group). NEC group was given hypoxia, cold stimulation and artificial feeding twice a day for 3 consecutive days. The control group was only fed normally. After modeling, From the 1st day to the 7th day, 10 rats were sampled in each group for pathological examination of colon, terminal ileum, lung, liver and kidney tissue. The levels of Bax, PCNA and PAF were investigated by immunohistochemistry. Results: Compared with the normal group, in the NEC group, on the 1st day, the colon, terminal ileum, lung, liver and kidney showed inflammatory damage. On the 5th day, the inflammatory injury was reduced. The inflammation disappeared on the 7th day. There were differences in the time of apoptosis in the intestine. In the intestine, the proliferation of PCNA was weak at first and then strong. Bax in liver and kidney showed marked apoptosis and apoptosis time increased in the lung. The expression of PCNA increased in lung, liver and kidney, and the expression of PAF increased in lung and liver. Conclusions: NEC can lead to secondary injury of different degrees in colon, terminal ileum, lung, liver and kidney, and the degree and time of injury and repair were different. In general, organ repair played a leading role on the 4th day after modeling.

Recent years have witnessed a rapid growth of distributed machine learning (ML) frameworks, which exploit the massive parallelism of computing clusters to expedite ML training. However, the proliferation of distributed ML frameworks also introduces many unique technical challenges in computing system design and optimization. In a networked computing cluster that supports a large number of training jobs, a key question is how to design efficient scheduling algorithms to allocate workers and parameter servers across different machines to minimize the overall training time. Toward this end, in this paper, we develop an online scheduling algorithm that jointly optimizes resource allocation and locality decisions. Our main contributions are three-fold: i) We develop a new analytical model that considers both resource allocation and locality; ii) Based on an equivalent reformulation and observations on the worker-parameter server locality configurations, we transform the problem into a mixed packing and covering integer program, which enables approximation algorithm design; iii) We propose a meticulously designed approximation algorithm based on randomized rounding and rigorously analyze its performance. Collectively, our results contribute to the state of the art of distributed ML system optimization and algorithm design.

Preterm birth and feeding are the most important pathogenic factors of neonatal necrotizing enterocolitis (NEC). Metabonomic has been widely used in the diagnosis and treatment of other diseases, but there is no research on the related diseases of premature infants. Compared with full-term infants, the metabolism of preterm infants has its own specificity, so it can easily lead to NEC and other digestive tract inflammatory diseases. Metabonomic may be applied to the diagnosis of preterm related diseases, such as NEC. The model was established with premature infant serum samples from 19 premature infants in our hospital, which was compared with the full-term infant control group. Serum was analyzed by gas chromatography-mass spectrometry (GC-MS), coupled with the analysis of serum metabolic characteristics. The variable important in projection, P value and Pearson correlation coefficient of samples were analyzed by using SIMCA, SPSS and other multivariate statistical analysis software. Compared to the term infants, premature infants had significantly higher levels of luteolin, and lower levels of xylose, O-succinyl-L-homoserine and lauric acid in the serum. There were some correlations among several different metabolites and clinically related indices (albumin, total bilirubin) for premature birth related diseases. There are metabolic alterations in the serum of premature infants, which make contribution to the diagnosis of NEC.

Deep Learning Recommendation Models (DLRMs) play a key role to power real-world recommendation and ranking, yet their growing complexity has made production deployment increasingly challenging. While PyTorch 2 (PT2) offers promising performance and productivity gains through automated model compilation, its initial release lacked critical features needed for DLRM adoption. In this work, we present RecCompl, a comprehensive model compilation system that enables efficient model compilation of industrial scale DLRMs with PT2. RecCompl addresses key compatibility issues by extending operator coverage, minimizing graph breaks, avoiding unnecessary recompilation, and generalizing graph transformation. Besides, to meet the high requirement on model exploration, we introduce a configuration-based user interface that decouples compilation settings from model code, allowing fine-grained control without intrusive changes. Despite these improvements, efficiency gaps remain in achieving a production-ready compilation system. To close them, we introduce systematic designs and engineering optimizations that enhance compilation time, memory management, and online deployment reliability. RecCompl is widely adopted, and delivers up to 60% higher training throughput while achieving substantially lower compilation latency, consistent performance across varying memory budgets, and stable online deployment.