Data Science & Analytics Reading List

This summer I got the opprotunity to mentor an intern in the Product Analytics team at Ancestry. I personally learned a lot from my mentor when I was an analyst intern three years ago (he later became my first full-time manager, and has been the one I turn to whenever I need career guidance). Therefore, I feel like it’s the perfect time for me to give back to those who helped me along the way. I am putting together this reading list as a gift to my intern, hoping except from the hands-on experience, he also gets more inspiration in his Data Science & Analytics career from this internship.

I. Product Sense & Analytics

Books

1. Lean Analytics
   Talks about metrics and analytics in the business setting
2. Web Analytics 2.0
   Focus on how to utilize web behavior data to understand user engagement and drive business insights

Readings

1. Tech Company DS&Analytics Blogs
   • Google
   • Linkedin
   • Facebook
   • Microsoft
   • Airbnb
   • Netflix
   • Uber
   • Lyft
   • Pinterest
2. Videos about Product Growth
   • Growth
   • How to get users and grow
3. Medium, especially the Towards Data Science site, is the place I go every week for short reading and inspiration. Below are some interesting articles about data science & analytics applications in the industry I recently read:
   • Data-Informed Product Building - A series of articles on how to use data to track and drive product growth
   • Product management & Analytics: what metrics should you be measuring - Discussed common metrics with different business context
   • How Instacart Uses Data Science to Tackle Complex Business Problems
   • Data Science in Walmart Supply Chain Technology
     
     

II. Experimentations

Online Courses

1. Udacity A/B Testing
   A very popular one on Udacity offered by Google. Detailed introduction on knowledge, application, and the whole process.

Books

1. Trustworthy Online Controlled Experiments: A Practical Guide to A/B Testing
   A/B tesing guiance with practical examples
2. Testing 1 - 2 - 3: Experimental Design with Applications in Marketing and Service Operations
   Focus more on technical details and knowledge behind multivariate testing, with applications in marketing filed mainly

Readings

1. A/B testing & Multivariate testing
   • 10 statistics traps in A/B testing – an ultimate guide for optimizers
     Common mistakes of implementing A/B testing
   • A/B and multivariate testing differences, advantages and limitations
     A/B test vs. multivariate test
2. Multi-armed bandits vs. A/B testing
   • A/B testing — Is there a better way? An exploration of multi-armed bandits
   • When and When Not to A/B Tests
3. Causal inference
   • Experimentation in Data Science
4. There are many other nuances around A/B testing, for example, network effect, testing with user opt-in, etc. Below are some helpful articles I came across. You can find more discussions online and in the tech company blogs above.
   • LinkedIn Feed Optimization - network effect, downstream/upstream metrics
   • Misadventure in Experiments for Growth - differences for A/B testing on established products vs. fledgling products
   • Compliance Bias in Mobile Experiments - utilize the propensity model when users’ experience does not always comply with the treatment given
   • Geo experimentation - run experiments on geo level when the network effect exists or when not being able to provide consistent user experience based on cookie
     
     

III. Machine Learning

Online Courses

There are just too many :) The one from Andrew Ng on Coursera is probably the most well-known one, but you will see 742 search results just with the keyword ‘Machine learning’ on Coursera. I personally enjoyed this Machine Learning specialization by the University of Washington when I was a beginner as it covers the most popular and basic algorithms with good business examples.

Books

1. An introduction to Statistical Learning
   Always known as ISLR. Intro textbook to most machine learning concepts and algorithms with R codes.
2. The Elements of Statistical Learning
   Always referred to as ESLR, another well-known textbook on machine learning.

Readings

Again, Medium is a good resource for short readings. Below are some articles I found helpful in recent months:

1. Classification
   • Pros and cons of various Machine Learning algorithms used in Classification
   • Decision Trees - understanding the basic of ensemble methods
   • Ensemble Methods – Bagging Boosting and Stacking
   • Catboost vs. LightGBM vs. XGboost
2. Clustering
   • Why So Many Different Clustering Algorithms?
3. Why How and When to Apply Feature Selection
4. The 5 most useful Techniques to Handle Imbalanced datasets
5. All about categorical variable encoding
6. Cross-validation the Right Way