本文介绍微软 AI 量化投资平台 Qlib 基础和进阶功能，对比传统量化策略 开发流程和 Qlib 提供的解决方案，提炼 Qlib 特色及优势，并探讨笔者使用 体会。Qlib 于 2020 年 9 月公开初版源码，2020 年 12 月获微软官网报道 并引发热议。我们认为 Qlib 的主要优势在于:1)覆盖量化投资全过程， 用户无需切换工具包或编程语言，降低 AI 算法使用门槛;2)从工程实现 角度，对因子数据储存、因子计算等环节提出创新解决方案，提升运算性 能和开发效率，或能解决量化投资研究中的部分痛点。

《量化投资以python为工具》主要讲解量化投资的思想和策略，并借助Python 语言进行实战。本书一共分为5部分，第1部分是Python 入门，第2部分是统计学基础，第3部分是金融理论、投资组合与量化选股，第4部分是时间序列简介与配对交易，第5部分是技术指标与量化投资。本书首先对Python 编程语言进行介绍，通过学习，读者可以迅速掌握用Python 语言处理数据的方法，并灵活运用Python 解决实际金融问题；其次，向读者介绍量化投资的理论知识，主要讲解量化投资所需的数量基础和类型等方面；最后讲述如何在Python语言中构建量化投资策略。

华泰证券人工智能系列1-41.rar

论文研究-ML-TEA：一套基于机器学习和技术分析的量化投资算法.pdf,  量化投资尝试利用计算机算法来预测证券的价格和进行证券的交易，并从中获取超额收益，是系统工程在金融投资领域的重要应用.本文设计了一套基于机器学习和技术指标的量化投资算法ML-TEA （machine learning and technical analysis）.该模型以技术指标作为输入变量，再分别通过不同的机器学习算法来预测股票数日之后的涨跌方向，并根据预测的方向来构建投资组合.实证结果显示：第一，三种模型的年化收益率都在25%以上，远超大盘指数的10.60%、买入持有策略的3%以及现有策略.从风险调节绩效（夏普比率、特雷纳比率和詹森绩效）来看，三种策略也都远超基准策略和现有策略.以夏普比率为例，三种策略均在1.50以上，而市场指数的夏普比率为0.38.第二，Ada-TEA和SVM-TEA都可以容忍远高于市场实际成本的交易成本.

After the fundamental volume and the advanced technique volume, this volume focuses on R applications in the quantitative investment area. Quantitative investment has been hot for some years, and there are more and more startups working on it, combined with many other internet communities and business models. R is widely used in this area, and can be a very powerful tool. The author introduces R applications with cases from his own startup, covering topics like portfolio optimization and risk management. There are six chapters in this book, categorized into three parts: Financial Market and Financial Theory, Data Processing and High Performance Computing of R, and Financial Strategy Practice. Every chapter is a holistic knowledge system. Section One is Financial Market and Financial Theory (including Chapters 1 and 2), which starts with an understanding of finance to establish a basic idea of financial quantification. Chapter 1, Financial Market Overview, is the opening chapter of this book, which mainly introduces the ideas and methods of how to use R language to make quantitative investments. Chapter 2, Financial Theory, mainly introduces the classic theoretical models of finance and the R implementation methods. In Section Two, Data Processing and High Performance Computing of R (including Chapters 3 and 4), essential tools of R language for data processing and their usage are introduced in detail. Chapter 3, Data Processing of R, cored with the data processing technology of R, introduces the methods of processing different types of data with R language. In Chapter 4, High Performance Computing of R, three external technologies are introduced to help the performance of R language meet the production environment requirements. Section Three, Financial Strategy Practice (including Chapters 5 and 6), combines the R language technology and the financial market rules to solve the practical problems in financial quantification field. In Chapter 5, Bonds and Repurchase, readers can learn the market and the methods of low-risk investment. In Chapter 6, Quantitative Investment Strategy Cases, the investment research methods from theory to practice are introduced in whole. Since knowledge of different areas is comprehensively applied in this book, it is suggested that you read all the chapters in order. Some of the technical implementations mentioned in this book use the information from the other two books of the series, R for Programmers: Mastering the Tools and R2 for Programmers: Advanced Techniques, so it is recommended that you read those two as well.

在机器学习中，遗传规划是一个优秀的特征生成工具，本文介绍了遗传规 划在 CTA 信号挖掘中的应用。遗传规划的优势在于能在现有的数据和运算 符中进行大规模启发式搜索，同时突破人类的思维局限，挖掘出特异的、 能对现有CTA策略做出有益补充的信号。本文基于遗传规划程序包gplearn 进行改进，并在多个商品期货品种上测试，挖掘出了一些回测效果较好且 意义明确的 CTA 信号。

给予证据分析的量化投资指南.

对量化投资的所涉及的知识方面、主要概念进行了梳理，建立了其间的层次结构图，便于爱好者按图索骥。

布林带突破策略（基于掘金客户端的python实现），本策略采用布林线进行均值回归交易。当价格触及布林线上轨的时候进行卖出，当触及下轨的时候，进行买入。

量化投资以python为工具 课后练习题代码 主要练习基本的量化策略和基本语法，学习python的基础函数和基本功能的实现