One of the recent best attempts at Text-to-SQL is pre-trained language model. Due to structural property SQL queries, seq2seq model takes responsibility parsing both schema items (i.e., tables and columns) skeleton keywords). Such coupled targets increase difficulty correct queries especially when they involve many logic operators. This paper proposes a ranking-enhanced encoding skeleton-aware decoding framework decouple linking parsing. Specifically, for encoder-decode model, its encoder...

Abstract With the tremendous success of machine learning (ML), concerns about their black-box nature have grown. The issue interpretability affects trust in ML systems and raises ethical such as algorithmic bias. In recent years, feature attribution explanation method based on Shapley value has become mainstream explainable artificial intelligence approach for explaining models. This paper provides a comprehensive overview value-based methods. We begin by outlining foundational theory rooted...

Abstract Naive Bayesian classification algorithm is widely used in big data analysis and other fields because of its simple fast structure. Aiming at the shortcomings naive Bayes algorithm, this paper uses feature weighting Laplace calibration to improve it, obtains improved algorithm. Through numerical simulation, it found that when sample size large, accuracy more than 99%, very stable; attribute less 400 number categories 24, 95%. empirical research, can greatly correct rate...

Pathological diagnosis of glioma subtypes is essential for treatment planning and prognosis. Standard histological based on postoperative hematoxylin eosin stained slides by neuropathologists. With advancing artificial intelligence (AI), the aim this study was to determine whether deep learning can be applied classification.

Lingxi Zhang, Jing Yanling Wang, Shulin Cao, Xinmei Huang, Cuiping Li, Hong Chen, Juanzi Li. Proceedings of the 61st Annual Meeting Association for Computational Linguistics (Volume 1: Long Papers). 2023.

In recent years, there have been a growing number of works studying the generalization properties stochastic gradient descent (SGD) from perspective algorithmic stability. However, few them devote to simultaneously and optimization for non-convex setting, especially pairwise SGD with heavy-tailed noise. This paper considers impact noise obeying sub-Weibull distribution on stability-based learning guarantees by investigating its jointly. Specifically, based two novel uniform model stability...

Disentangled Representation Learning (DRL) aims to learn a model capable of identifying and disentangling the underlying factors hidden in observable data representation form. The process separating variation into variables with semantic meaning benefits learning explainable representations data, which imitates meaningful understanding humans when observing an object or relation. As general strategy, DRL has demonstrated its power improving explainability, controlability, robustness, as well...

Neutrophil-to-lymphocyte ratio (NLR), as a novel inflammatory marker, has been shown to be associated with the severity and prognosis of various cardiovascular diseases. The aim this study was investigate whether NLR can serve biomarker for adverse outcomes prognostic value in patients dilated cardiomyopathy (DCM). This retrospective analysis 666 consecutive DCM who were admitted our center first time. We compared levels among different outcome groups assessed survival status categories....

Few-shot knowledge graph completion is to infer the unknown facts (i.e., query head-tail entity pairs) of a given relation with only few observed reference pairs. Its general process first encode implicit an pair and then match relations Most existing methods have thus far encoded matched pairs by using direct neighbors concerned entities. In this paper, we propose P-INT model for effective few-shot completion. First, infers leverages paths that can expressively two Second, capture fine...

Adaptive feature extraction is useful in many information processing systems. This paper proposes a learning machine implemented via neural network to perform such task using the tool principal component analysis. (1) adaptive nonstationary input, (2) based on an unsupervised concept and requires no knowledge of if, or when, input changes statistically, (3) performs online computation that little memory data storage. Associated with this machine, authors propose algorithm (LEAP), whose...

As the traffic flow has characteristics of non-linear and strong interference, it different features in time-frequency domain. The traditional short-term forecasting methods have disadvantages lower prediction accuracy, harder parameter determination poorer adaptability. Aiming at above problems, we proposed a short - term algorithm based on wavelet function Extreme Learning Machine (ELM) to optimize method. Firstly, activation hidden layer neurons model ELM is optimized according denoising...

Multi-agent deep reinforcement learning (MADRL) has attracted a tremendous amount of interest in recent years. In this paper, we introduce MADRL into the confrontation scene Unmanned Aerial Vehicle (UAV) swarm. To analysis dynamic game process UAV swarm confrontation, build two non-cooperative models based on paradigm. By using multi-agent deterministic policy gradient (MADDPG) and centralized training with decentralized execution method, achieve Nash equilibrium under 5 vs. scenes. We also...

Existing methods on knowledge base question generation (KBQG) learn a one-size-fits-all model by training together all subgraphs without distinguishing the diverse semantics of subgraphs. In this work, we show that making use past experience semantically similar can reduce learning difficulty and promote performance KBQG models. To achieve this, propose novel approach to with meta-learner (DSM). Specifically, devise graph contrastive learning-based retriever identify subgraphs, so construct...

Decentralized Stochastic Gradient Descent (D-SGD) represents an efficient communication approach tailored for mastering insights from vast, distributed datasets. Inspired by parallel optimization paradigms, the incorporation of minibatch serves to diminish variance, consequently expediting process. Nevertheless, as per our current understanding, existing literature has not thoroughly explored learning theory foundation Minibatch (DM-SGD). In this paper, we try address theoretical gap...

D-dimer is a biomarker of coagulation and fibrinolytic system activation in response to the hypercoagulable state body. The research aimed analyze value prognosis patients with dilated cardiomyopathy (DCM). Patients admitted our center for first time DCM were enrolled consecutively. clinical characteristics variables obtained from electronic medical record system, prognostic information was using telephone return visits review repeated hospitalization records. Univariate multivariate Cox...

Knowledge Base Question Answering (KBQA) is to answer natural language questions posed over knowledge bases (KBs). This paper targets at empowering the IR-based KBQA models with ability of numerical reasoning for answering ordinal constrained questions. A major challenge lack explicit annotations about properties. To address this challenge, we propose a pretraining model consisting NumGNN and NumTransformer, guided by self-supervision signals. The two modules are pretrained encode magnitude...

The recent rise of conversational applications such as online customer service systems and intelligent personal assistants has promoted the development knowledge base question answering (ConvKBQA). Different from traditional single-turn KBQA, ConvKBQA usually explores multi-turn questions around a topic, where ellipsis coreference pose great challenges to KBQA which require self-contained questions. In this paper, we propose rewrite-and-reason framework first produce full-fledged rewritten...

The development of neural machine translation has achieved a good effect on large-scale general corpora, but there are still many problems in the low resources and specific fields. This paper studies problem field electrical engineering fuses multi-layer vectors at encoder side model. On this basis, decoder unit model is improved, multi-attention mechanism based vector fusion proposed, which improves ability to extract features achieves better Chinese-English tasks. experimental results show...