This book is perfect to get you started with probabilistic graphical models (PGM) with Python. It starts with a quick intro to Bayesian and Markov Networks covering concepts like conditional independence and D-separation. It then covers the different aspects of PGM: structure learning, parameter estimation (with frequentist or Bayesian approach) and inference. All is illustrated with examples and code snippets using mostly the libpgm package. PyMC is used for Bayesian parameter estimation.

There are two approaches to undergraduate and graduate courses in linear statistical models and experimental design in applied statistics. One is a two-term sequence focusing on regression followed by ANOVA/Experimental design. Applied Linear Statistical Models serves that market. It is offered in business, economics, statistics, industrial engineering, public health, medicine, and psychology departments in four-year colleges and universities, and graduate schools. Applied Linear Statistical Models is the leading text in the market. It is noted for its quality and clarity, and its authorship is first-rate. The approach used in the text is an applied one, with an emphasis on understanding of concepts and exposition by means of examples. Sufficient theoretical foundations are provided so that applications of regression analysis can be carried out comfortably. The fourth edition has been updated to keep it current with important new developments in regression analysis.

THULAC，全称为“清华树洞分词系统”（Tsinghua Lexical Analysis System），是由清华大学自然语言处理与社会人文计算实验室开发的一款中文分词和词性标注工具。这个工具包，`thulac.models`，是THULAC的核心模型部分，用于执行高效的中文文本处理任务。 我们要理解什么是分词。在自然语言处理中，分词是将连续的文本序列分割成具有独立语义的词语单元，这是进行后续文本分析如词性标注、情感分析、命名实体识别等的基础步骤。THULAC的优势在于它结合了基于词典的分词方法和基于统计的分词模型，既考虑了词语的规范性，又适应了语言的灵活性，特别适合处理现代汉语的各种复杂情况。 THULAC的工作流程大致分为以下几个步骤： 1. **加载模型**：`thulac.models`中的模型包含了大量训练数据得到的参数，用于识别和划分词语。在使用THULAC时，我们需要先加载这些模型。 2. **预处理**：对输入的文本进行预处理，如去除标点符号、数字等非汉字字符。 3. **分词**：使用加载的模型对预处理后的文本进行分词。THULAC支持两种模式：精细模式和粗略模式。精细模式更注重词语的准确性，而粗略模式则更注重速度。 4. **词性标注**：THULAC不仅分词，还会为每个词标注其对应的词性，这有助于进一步的语义理解和分析。 5. **后处理**：根据需求，可以对分词和词性标注的结果进行清洗和整理。 在实际应用中，`thulac.models`通常与其他编程语言库（如Python的`thulac`库）配合使用。用户只需按照库的API调用模型，即可方便地实现分词和词性标注功能。例如，在Python中，你可以通过以下代码进行操作： ```python from thulac import Thulac thu = Thulac(seg_only=True) # 使用默认设置，仅进行分词 words = thu.cut('这是一个分词示例。') for word in words: print(word) ``` 标签中的"清华分词工具包 THULAC thulac.models"，表明了`thulac.models`是THULAC工具包的一部分，主要包含模型文件，而"thulac"可能是指整个THULAC的Python实现库。 `thulac.models`是THULAC分词工具的核心组件，提供强大的中文分词和词性标注能力，广泛应用于学术研究、新闻分析、社交媒体监控等各种领域。对于处理中文文本的自然语言处理项目来说，这是一个非常有价值的资源。

从DDPM到score-based generative models再到Consistency Models的介绍，对于扩散模型的全面理解有一定的帮助。

用于无法访问hugging face并需要运行stable-diffusion-webui时使用

基于广义加乘危险率模型的多元数据病例队列试验研究，郑明，孙怡，在大型流行病学跟踪研究中,病例队列抽样设计是一种常用的降低抽样成本的抽样方法。在这种抽样设计中,只要求观测所有病例样本以及�

2024最新Sora论文（Video generation models as world simulators）的引用文件，包含bib、txt（用于endnote），rdf。 - 方便论文写作直接使用

Petri Nets Fundamental Models

【多模态大模型综述】 使用 gpt3.5 精细翻译，完美融合图片等内容 由微软7位华人研究员撰写，足足119页 它从目前已经完善的和还处于最前沿的两类多模态大模型研究方向出发，全面总结了五个具体研究主题： - 视觉理解 - 视觉生成 - 统一视觉模型 - LLM加持的多模态大模型 - 多模态 agent 本报告一共7位作者。 发起人和整体负责人为 Chunyuan Li。 他是微软雷德蒙德首席研究员，博士毕业于杜克大学，最近研究兴趣为 CV 和 NLP 中的大规模预训练。 他负责了开头介绍和结尾总结以及“利用 LLM 训练的多模态大模型”这章的撰写。 核心作者一共 4位： Zhe Gan Zhengyuan Yang Jianwei Yang Linjie Li 他们分别负责了剩下四个主题章节的撰写。

Ross's classic bestseller, Introduction to Probability Models, has been used extensively by professionals and as the primary text for a first undergraduate course in applied probability. It provides an introduction to elementary probability theory and stochastic processes, and shows how probability theory can be applied to the study of phenomena in fields such as engineering, computer science, management science, the physical and social sciences, and operations research. With the addition of several new sections relating to actuaries, this text is highly recommended by the Society of Actuaries. A new section (3.7) on COMPOUND RANDOM VARIABLES, that can be used to establish a recursive formula for computing probability mass functions for a variety of common compounding distributions. A new section (4.11) on HIDDDEN MARKOV CHAINS, including the forward and backward approaches for computing the joint probability mass function of the signals, as well as the Viterbi algorithm for determining the most likely sequence of states. Simplified Approach for Analyzing Nonhomogeneous Poisson processes Additional results on queues relating to the (a) conditional distribution of the number found by an M/M/1 arrival who spends a time t in the system,; (b) inspection paradox for M/M/1 queues (c) M/G/1 queue with server breakdown Many new examples and exercises.