|
所在平台: Udemy |
课程主页: https://www.udemy.com/course/optimization/
课程评论：没有评论
**课程名称：** 优化与 A/B 测试统计学 **课程概述：** 本课程旨在帮助您掌握优化和 A/B 测试的核心知识，加速业务增长。无论您是在初创公司还是大型企业，本课程都将为您提升技能。课程涵盖 A/B 测试的基础知识、运行 A/B 测试的 8 个关键步骤，并深入讲解假设检验背后的统计学原理。通过正确设置测试和严谨的统计分析，您可以为组织节省宝贵的时间和精力，并可能带来数倍的投资回报。 **课程内容亮点：** * A/B 测试的应用案例 * 假设检验 * 将测量作为降低风险的手段 * 选择关键绩效指标 (KPI) 或成功指标 * 运行 A/B 测试的 8 个步骤 * A/B 测试与多变量测试 (MVT) 设计的选择 * 提升阈值 (Lift Threshold) * 零假设 (Null Hypothesis) * 统计显著性 (Statistical Significance) * 样本量估算 * 置信区间 (Confidence Interval) * 检验统计量 (Test Statistic) * t 检验 (t-tests) * 均值标准误 (Standard Error of the Mean) * 卡方检验 (Chi-square) * 费舍尔精确检验 (Fischer Exact test) * 统计功效 (Statistical Power) * 第一类错误 (Type I error) * 第二类错误 (Type II error) * p 值 (p-values) * 如何选择合适的统计检验方法 **学习收益：** * 掌握运行可靠 A/B 测试的完整流程。 * 理解 A/B 测试背后的统计学原理，能够进行严谨的数据分析。 * 学会如何选择合适的指标和统计方法，避免常见错误。 * 提升数据驱动决策的能力，为组织带来实际价值。 * 在工作中展现专业能力，令同事印象深刻。
Whether you've got a lean startup or a fat Fortune 500, the faster you learn the faster you'll grow. Optimization and a/b testing is at the heart of learning fast. I guarantee you will learn something in this course that will raise your skill level. With the 30-day money-back guarantee, you can't lose. We start with the basics, then cover the 8 steps of running a solid a/b test. Next we dive deep into the statistics behind hypothesis testing. In the long-run you will save your organization headaches by setting up tests correctly and analyzing them with the right statistical rigour. There is double and triple digit ROI around optimization for companies that figure it out. Start now and impress your colleagues on Monday morning. Topics include: Examples of a/b tests Hypothesis testing Measurement as risk reduction Selecting a KPI or success metric 8 Steps for Running an A/B Test Selecting from amongs a/b test and MVT test designs Lift Threshold Null Hypothesis Statistical significance Sample size estimates confidence interval test statistic t-tests standard error of the mean chi-square Fischer Exact test Statistical Power Type I error Type II error p-values How to choose what statistical test to run
