本期广证恒生新三板教育指数上涨0.66%，领先三板成分指数1.24 个百分点，领先三板做市指数2.11 个百分点。本期新增新金宝威、环球优学2 家挂牌企业。

新三板教育行业正从超高速增长逐步回落至较快速增长新常态，业绩分化加剧，强者恒强，建议投资者结合当前确定性较高的Pre-IPO 投资策略，聚焦质优成长股。

商誉。实践中，商誉=购买企业支付对价的公允-被购买企业的可辨认净资产的公允，商誉一般产生于非同一控制下的企业合并。

目前好未来市值96 亿，新东方市值102 亿，K12 教培业务内生增长强劲带动“双巨头”股价屡创新高。报名人数的增长是好未来和新东方K12 教培收入的增长的主要动力，ASP 增速有限。好未来在核心城市仍能实现翻倍的收入增长，体现强大的品牌效应，未来双师课堂在核心城市的拓展可进一步满足旺盛的需求。

截至 2017 年 4 月 17 日，随着教育行业整体的高速增长，行业 整体估值水平逐渐降低。总体来看新三板教育行业为28.11倍PE,。 A股教育行业估值较高为41.37倍PE。值得注意的是A股市场PRE-K 板块利润增速较快。新三板留学行业估值高，利润增速也快，职业 以及幼教利润增速明显且估值较低。目前，我们看好 A 股幼教行 业、新三板幼教以及职业教育细分板块，随着市场进一步企稳， 2017 年婴儿潮的爆发，教育行业有望站上风口，建议重点关注教 育板块。

龙门教育 2016 年净利润同比增长 44%。2017 年 4 月 12 日，龙门教育公告 2016 年年报，全年实现收入 2.4 亿元，同比增长 60%；归 母净利润 7,029 万元，同比增长 44%。（ 1）西安全封闭校区(中高补、高 一高二)培训学生同比增加 5,809 人，培训费收入增加 5,504 万元，高三 重点班一本上线率 53%，二本上线率 94%，中考部上线率 92%，规模和 教学质量同步提升；（2）新增 K12 课外辅导业务，从区域转战全国，并 实现 O2O 远程直播互动教学，同期培训学生超过 500 人，已覆盖 12 个 城市，开设 56 个校区，增收 2,100 万元；（ 3）多元化销售渠道布局驱动 教辅软件业务增收 1,319 万元

In the past two decades, globalization (and technological advances) has changed the dynamics of the education market and the corporate training market. The emergence of new types of personal computing devices - coupled with the introduction of technologies (such as gamification, apps, and social networks) - has upgraded learning and development opportunities for employees significantly.

In the past two decades, globalization combined with technological advances has changed the dynamics of the education and corporate training markets worldwide. The emergence of new types of personal computing devices coupled with technologies such as gamification, apps, and social networks has enhanced the learning and development opportunities for employees significantly.

Gamification is the process of implementing game mechanics into non-gaming contexts to drive user engagement and to enhance problem-solving. Points, badges, leaderboards, challenges, and rewards are some examples of game mechanics. Gamification does not create real games but uses game techniques to engage students in comprehensive learning mechanisms.

Deep learning is the subset of machine learning and a method of training a system by itself. Machine learning is a software or technique that improves the performance of the system over time when determined against a specified task. It stems from artificial neural networks (ANNs). A neural network is a massively parallel distributed mainframe made up of simple processing units and has a natural tendency for storing observed knowledge and making it available to the user. The following diagram shows the subsets AI to provide a clear picture of the relationship between deep learning and machine learning.