Data Analytics under the Business View

Introduction to the data analytics methods hinged with the business thinking.

Course Overview

With the explosion of Deep Learning, AI(Artifacial Intelligence), ML (Machine Learning) and Deep Learning etc. are becoming more and more hot in recent years. This course tries to cover those related topics in an interesting way which adopts the business thinking as the framework.

To do so, this course provides a big picture as follows:

  • Overview of Business Thinking (simplified MBA)
  • Evolution of AI (clarify the relationship for AI, ML/DM, DL etc.)
  • A little Optimization (a key to understand the related ML algorithms)
  • Data Analytics methods/algorithms from Statistics and ML
  • Practical skills in real big e-commerce (to cater for "秒杀" and "Precision adv.")

Schedule

Lectures Description Course Materials
1 Introduction Why and how do I demonstrate Data Analytics methods under Business view ?
2 So called business
  • 3 element model to undertand business activities
  • 3 chain analysis - Customer chain, Value chain, Production Chain
  • Equations to evaluate the profitability
[02MBA-So called business]
3 Management Skills
  • Strategy Planning: (SWOT) – Cost analysis is important!
  • Find the right guy (DISC,MBTI), …
  • Ensure the executive force:
    • Balanced Score Board + KPI,
    • Why-Why, Fishbone, Pareto, …
    • Brainstorm, Osborn’s Checklist, …
 [03MBA-Management Skills]
4 History of Data Analytics It's a long way to pursue wisdom:
  • Before IT, Math is the crown, and human is the actor
  • In IT age, AI is the symbol with 3 stages
    - Logic Inference, Knowledge Engineering, Machine Learning
  • Future is still on its road
[04History of DA-pursue wisdom]
5 A little Optimziation Just the Extreme Value theorem :) [05Optimization-the key to understand DA]
6 You should learn Statistics Many DA methods are from Statistics [06]
7-A Basic ideas before ML Ideas about Distance, Manifold, PCA, Kernel, EM etc. [07ML basic A-Ideas-Distance, Spaces and Hidden Parameters(SPOC)]
7-B Association Rule A Priori algorithm [07-B]
7-C Clustering without GMM Clustering without GMM like K-means, and DBScan [7-C]
7-D Classification without SVM Decision tree, Logistics Regression, Assemble algorithm like Ada-boost [7-D]
7-E Neural Network Perceptron, Logistics Regression, MLP, Hopfield NN, etc. [7-E]
8-A DL DL like CNN, LSTM, GAN etc. [8-A]
8-B SVM SVM - Support Vector Machine [8-B]
8-C MDP MDP - Markov Decision Process [8-C]
8-D GMM GMM - Gaussian Mixture Model [8-D]
8-E Topic Modeling pLSA, LDA, Peacock etc. [8-E]
9 Big Analytics, Big Computing Big Analytics, Big Computing [09Big Analytics, Big Computing]
10 Recommendation & Computing Adv. Recommendation & Computing Adv. [10Recommendation & Computing Adv]

Course Logistics and Policies

Grading In-class quizzes: 50% (5 times), Exam: 50%.

Textbook There is no required textbook, but for students who want additional resources, we recommend the following two:
  • 朱燕云."褚时健: 影响企业家的企业家". 2014年11月1日
  • 周桦. 褚时健传. 中信出版集团, 2015.
  • 尼格尼维斯基. 人工智能: 人工智能·智能系统指南（原书第3版）.机械工业出版社, 2012.
  • Stephen Boyd and Lieven Vandenberghe. Convex Optimization. Cambridge University Press.
  • 李航. 统计学习方法（第2版）. 清华大学出版社, 2019-5-1.
  • Pang-Ning Tan, Michael Steinbach, Vipin Kumar. Introduction to Data Mining. 2013.
  • Ian Goodfellow, Yoshua Bengio, Aaron Courville. Deep Learning.
  • 刘鹏, 王超 . 计算广告. 人民邮电出版社. 2015.