项目方案讨论会流程_讨论本文可能的解决方案

项目方案讨论会流程

The text passage must include use-case description, problem description, identifiable features that may lead to an Agile Storyboard. Each text passage is allocated specific value for Story Points. From the Story Board, features are extracted into Subject Area, Features Sets and Features.

文字段落必须包括用例描述，问题描述，可能导致敏捷情节提要的可识别功能。 每个文本段落都被分配了故事点的特定值。 从情节提要中，要素被提取到主题区域，要素集和要素中。

敏捷需求精炼厂(ARR) (Agile Requirements Refinery (ARR))

(1) Fit Criterion

(1)适合标准

(2) Traceability

(2)可追溯性

(3) Refinery

(3)炼油厂

(4) Deployment

(4)部署

(4)部署的ARR (ARR for (4) Deployment)

Separating configuration of project from the project itself. Making the configuration sit on a separate location and deploying the project from an IDE that manages project configuration such as composer.json / POM, etc.

将项目的配置与项目本身分开。 使配置位于不同的位置，并从管理项目配置的IDE(例如composer.json / POM等)部署项目。

(2)可追溯性的ARR (ARR for (2) Traceability)

The Dotted notations shown on accordion, which shows the status of the requirements. If it is a subject area then it shows number of feature sets, if a feature, it shows the requirement description in single liner.

在手风琴上显示了点符号，它显示了需求的状态。 如果是主题区域，则显示多个功能集；如果为特征，则显示单个衬里中的需求描述。

(3)炼油厂的ARR (ARR for (3) Refinery)

Patch Management using Refinery for the Sprint. For example, managing logic for composer.json, such as specifications for version usages, by using git-svn.

使用精炼厂进行Sprint的补丁程序管理。 例如，通过使用git-svn管理composer.json的逻辑，例如版本用法的规范。

(1)适合标准的ARR (ARR for (1) Fit Criterion)

Setup validation constraints for the substring off requirement for every fit criterion requirement represented as an FDD.

为每个拟合标准要求的子串关闭要求设置验证约束，以FDD表示。

(2)可追溯性的ARR (ARR for (2) Traceability)

Search through the requirements from child to parent and vice-versa.

搜索从孩子到父母的要求，反之亦然。

(1)适合标准的ARR (ARR for (1) Fit Criterion)

The project blastoff stage will expose the metrics discussed in the first stakeholder meeting such as accuracy, benchmark, rest interfaces, implementation roadmap, estimated cluster points, cluster points cascaded and staged similar to a Gantt Chart.

项目启动阶段将公开在第一次利益相关者会议上讨论的度量标准，例如准确性，基准，休息界面，实施路线图，估计的聚类点，与甘特图类似的级联和上演的聚类点。

(3)炼油厂的ARR (ARR for (3) Refinery)

The cluster of requirements are categorized into a distance metric based on different AST based parsing algorithms. The DEA data is obtained by a covariance matrix using Word2Vec and the results of AST parsing. A pre-configured AST based parsers are made available and the document routed to golden record set.

基于不同的基于AST的解析算法，需求集群被分类为距离度量。 通过使用Word2Vec的协方差矩阵和AST解析的结果获得DEA数据。 提供了一个预配置的基于AST的解析器，并将文档路由到黄金记录集。

(1)适合标准的ARR (ARR for (1) Fit Criterion)

Convert the DEA problem to an inertia bias problem using the given DMU(s) using a Gaussian from Word2Vec and outlier theory. This will size the problem further based on applications of decision theory. This will ensure the Gantt Chart cascading is performed properly.

使用Word2Vec中的高斯和离群值理论，使用给定的DMU将DEA问题转换为惯性偏差问题。 根据决策理论的应用，这将进一步解决问题。 这将确保正确执行甘特图级联。

I have a reason to believe this because Cognitive Representations are Systematic and Compositional. So if we adhere to the sample AST parsed requirements, then that would be confirming the bias problem. Once an estimated Gantt Chart is done, it is easy to plan for the later stages.

我有理由相信这一点，因为认知表示是系统的和组合的。 因此，如果我们遵守样本AST解析的要求，那么这将证实存在偏差问题。 一旦完成了估计的甘特图，很容易为以后的阶段进行计划。

(2)可追溯性的ARR (ARR for (2) Traceability)

Decision for GO/NO-GO is made based on passes through Refinery on based on pre-computed metrics such as Gini Impurity and Min. Mean Square Estimation.

是否通过/不通过的决定是基于通过精炼厂的流量，并基于预先计算的指标(例如基尼杂质和最小值)进行的。 均方估计。

(4)部署的ARR (ARR for (4) Deployment)

This deals with local environment based deployments, such as shell scripts, that run and execute a Docker box for example.

这涉及基于本地环境的部署，例如运行和执行Docker box的Shell脚本。

(1)炼油厂的ARR (ARR for (1) Refinery)

The inertia bias problem is approximated with pi, to find the new DMU(s) to solve the stewardship problem. Each AST tree is measured using depth vs breadth, as well as the area it covers, the tree depth is decided based on resolution of requirements and the tree split is decided based on dimensional information of the representation problem such as subject area, feature sets and features will have a split size of 2.

用pi近似惯性偏差问题，以找到新的DMU来解决管理问题。 使用深度与宽度以及覆盖的面积来测量每棵AST树，根据需求的分辨率确定树的深度，并根据表示问题的维度信息(例如主题区域，特征集和功能的分割尺寸为2。

Aswin Vijayakumar

阿斯温·维杰亚库玛(Aswin Vijayakumar)

翻译自: https://medium.com/@aswinkvj/discussion-of-possible-solutions-for-the-article-a6e651ad70e5

项目方案讨论会流程