Computer Vision Metrics provides an extensive survey and analysis of over 100 current and historical feature description and machine vision methods, with a detailed taxonomy for local, regional and global features. This book provides necessary background to develop intuition about why interest point detectors and feature descriptors actually work, how they are designed, with observations about tuning the methods for achieving robustness and invariance targets for specific applications. The survey is broader than it is deep, with over 540 references provided to dig deeper. The taxonomy includes search methods, spectra components, descriptor representation, shape, distance functions, accuracy, efficiency, robustness and invariance attributes, and more. Rather than providing ‘how-to’ source code examples and shortcuts, this book provides a counterpoint discussion to the many fine opencv community source code resources available for hands-on practitioners.

一个从开源项目 综合数据生成项目的指标 网址： ： 文档： : 仓库： : 执照： 发展状况： 概述 SDMetrics库提供了一组与数据集无关的工具，用于通过将综合数据库与建模后的真实数据库进行比较来评估综合数据库的质量。 它支持多种数据模式： 单列：比较代表各个列的一维numpy数组。 列对：比较pandas.DataFrame列如何pandas.DataFrame关联（以2组为一组）。 单个表：比较整个表，以pandas.DataFrame表示。 多表：将以python dict表示的多表和关系数据集与以pandas.DataFrame传递的多个表进行pandas.DataFrame 。 时间序列：比较代表事件顺序的表格。 它包括各种指标，例如： 使用统计检验比较实际和合成分布的统计量度。 使用机器学习来尝试区分真实数据和合成数据的检测指标。 效能

CNNIQA 以下论文的PyTorch 1.3实施： 笔记 在这里，选择优化器作为Adam，而不是本文中带有势头的SGD。 data /中的mat文件是从数据集中提取的信息以及有关火车/ val /测试段的索引信息。 LIVE的主观评分来自。 训练 CUDA_VISIBLE_DEVICES=0 python main.py --exp_id=0 --database=LIVE 训练前， im_dir在config.yaml被指定必须的。 可以在config.yaml设置数据库内实验中的Train / Val / Test拆分比率（默认值为0.6 / 0.2 / 0.2）。 评估 测试演示 python test_demo.py --im_path=data/I03_01_1.bmp 交叉数据集 python test_cross_dataset.py --help TODO：

DesigniteJava DesigniteJava是用于用Java编写的代码的代码质量评估工具。 它检测到大量的设计和实现气味。 它还计算许多常用的面向对象的指标。 特征 检测出17种设计气味 命令式抽象 多面抽象 不必要的抽象 未利用的抽象 封装不足 未利用的封装 模块化破损 循环依赖的模块化 模块化不足 集线器式模块化 层次结构破裂 循环层次 深层次 缺少层次结构 多路径层次结构 叛逆阶层 广泛的阶层 检测10种实施气味 构造函数的抽象函数调用 复杂条件 复杂方法 空捕获子句 长标识符 长方法 长参数列表 长声明 幻数 缺少默认值 计算以下面向对象的指标 LOC（代码行-方法和类的粒度） CC（圈复杂度-方法） PC（参数计数-方法） NOF（字段数-类） NOPF（公共字段数-类） NOM（方法数量-类） NOPM（公共方法数量-类） WMC（每类的加权方法-类）

博客https://blog.csdn.net/lsshlsw/article/details/82670508 spark_prometheus_metrics.json

不确定性工具箱 用于预测不确定性量化，校准，python工具箱。 另外：的以及的集合。 许多机器学习方法会返回预测以及某种形式的不确定性，例如分布或置信区间。 这就引出了一个问题：我们如何确定最佳的预测不确定性？ 产生最佳或理想不确定性是什么意思？ 我们的不确定性是否准确且经过良好校准？ 不确定性工具箱提供了用于量化和比较预测性不确定性估计值的标准度量，提供了这些度量的直觉，生成了这些度量/不确定性的可视化效果，并实施了简单的“重新校准”程序来改善这些不确定性。 该工具箱当前专注于回归任务。 工具箱内容 不确定性工具箱包含： 与预测不确定性量化相关的。 评估预测不确定性估计的质量的。 预测不确定性估计和指标。 用于改善预测变量校准的方法。 有关度量标准和方法的相关。 安装 不确定性工具箱需要Python 3.6及更高版本。 要安装，请克隆并通过cd进入此仓库，然后运行： $

APM理解 APM（Application Performance Management/Monitor）通常包含三个分支：指标（Metrics）、追踪（Tracing）、日志（Logging），三者之间的关系[4]描述很到位。简单说Metrics是时序性的数值指标，通常用于服务监控；Tracing针对请求链，对理解复杂应用的拓扑结构，发现性能问题很有用；日志通常记录连续的事件信息。三个分支相互交融，业界的APM系统分别有所侧重。 Metrics是传统系统监控手段，出现的比较早，典型的有Prometheus[5]、Ganglia[6] Tracing随着大型分布式的出现而兴起，以google的

“When will it be done?” That is probably the first question your customers ask you once you start working on something for them. Think about how many times you have been asked that question. How many times have you ever actually been right? We can debate all we want whether this is a fair question to ask given the tremendous amount of uncertainty in knowledge work, but the truth of the matter is that our customers are going to inquire about completion time whether we like it or not. Which means we need to come up with an accurate way to answer them. The problem is that the forecasting tools that we currently utilize have made us ill-equipped to provide accurate answers to reasonable customer questions. Until now. Topics Include Why managing for flow is the best strategy for predictability—including an introduction to Little’s Law and its implications for flow. A definition of the basic metrics of flow and how to properly visualize those metrics in analytics like Cumulative Flow Diagrams and Scatterplots. Why your process policies are the potentially the biggest reason that you are unpredictable. Table of Contents PART ONE - FLOW FOR PREDICTABILITY Chapter 1 - Flow, Flow Metrics, and Predictability Chapter 2 - The Basic Metrics of Flow Chapter 3 - Introduction to Little’s Law PART TWO - CUMULATIVE FLOW DIAGRAMS FOR PREDICTABILITY Chapter 4 - Introduction to CFDs Chapter 5 - Flow Metrics and CFDs Chapter 6 - Interpreting CFDs Chapter 7 - Conservation of Flow Part I Chapter 8 - Conservation of Flow Part II Chapter 9 - Flow Debt PART THREE - CYCLE TIME SCATTERPLOTS FOR PREDICTABILITY Chapter 10 - Introduction to Cycle Time Scatterplots Chapter 10a - Cycle Time Histograms Chapter 11 - Interpreting Cycle Time Scatterplots Chapter 12 - Service Level Agreements PART FOUR - PUTTING IT ALL TOGETHER FOR PREDICTABILITY Chapter 13 - Pull Policies Chapter 14 - Introduction to Forecasting Chapter 15 - Monte Carlo Method Introduction Chapter 16 - Getting Started PART FIVE - A

Project tracking systems, test and build tools, source control, continuous integration, and other built-in parts of the software development lifecycle generate a wealth of data that can be used to track and improve the quality and performance of products, processes, and teams. Although the iterative nature of Agile development is perfect for data-driven continuous improvement, the collection, analysis, and application of meaningful metrics often fades in favor of subjective measures that offer less insight into the real challenges of making better software. Agile Metrics in Action: Measuring and enhancing the performance of Agile teams is a practical book that shows how to take the data already being generated to make teams, processes, and products better. It points out which metrics to use to objectively measure performance and what data really counts, along with where to find it, how to get it, and how to analyze it. The book also shows how all team members can publish their own metrics through dashboards and radiators, taking charge of communicating performance and individual accountability. Along the way, it offers practical data analysis techniques, including a few emerging Big Data practices. Purchase of the print book includes a free eBook in PDF, Kindle, and ePub formats from Manning Publications. Table of Contents Part 1 Measuring agile teams Chapter 1 Measuring agile performance Chapter 2 Observing a live project Part 2 Collecting and analyzing your team’s data Chapter 3 Trends and data from project-tracking systems Chapter 4 Trends and data from source control Chapter 5 Trends and data from CI and deployment servers Chapter 6 Data from your production systems Part 3 Applying metrics to your teams, processes, and software Chapter 7 Working with the data you’re collecting: the sum of the parts Chapter 8 Measuring the technical quality of your software Chapter 9 Publishing metrics Chapter 10 Measuring your team against the agile principles Appendix A D

metrics-server/metrics-server:v0.5.0 已打包 docker load -i metric-serverv0.5.0.tar && docker image ls