The Human Learning of Machine Learning: a reading list for Executives, Managers and those wanting to go the distance
Analytics is an accelerating field. And no doubt, you often enough hear its latest and greatest jargon-innovation flung around like common punctuation: Big Data, Predictive Analytics, Data mining, Machine Learnings, Deep Learning, NLP, Artificial Intelligence etc. etc.
As you weed through, if you remain awe-struck by the genuine depth of insight or the outlandish predictions this new science can make, you might crave some genuine understanding... but where to start?
Let me share a few resources to help. There’s 3 levels of depth you can go to:
1)You want to understand the growing applications in business, government or science, and the level of potential impact - e.g., what are inspirational analytics-driven companies doing?
2)You want to understand different techniques and algorithm approaches - e.g., what does machine learning really mean? How does it work?
●3)You want to be able to run analytics techniques yourself and build predictive models - e.g., what do I need to build my own predictive models?
Decide how deep you want to go and here's my view of best resources to get started (open to hearing additions):
1) The growing use of analytics, data science and AI:
A. Predictive Analytics by Eric Siegel https://www.amazon.in/Predictive-Analytics-Power-Predict-Click-ebook/dp/B019HR9X4U?_encoding=UTF8&btkr=1&ref_=dp-kindle-redirect
-> good explanations of how analytics has driven the success of several companies: Neflix, Chase Bank, Telenor, the Obama campaign! Also, to understand how IBM's Watson really works
B. Big Data in Practice: How 45 Successful Companies Used Big Data to Deliver Extraordinary Results by Bernard Marr http://www.amazon.in/Big-Data-Practice-Successful-Extraordinary/dp/8126562811/ref=sr_1_1?s=books&ie=UTF8&qid=1486030242&sr=1-1&keywords=bernard+marr
-> good read to be amazed by how non-tech companies like Walmart, Rolls Royce, Ralph Lauren, Shell, the BBC are actually companies where analytics is already fundamental to how they tick
C. McKinsey Global Institute's latest 'Age of Analytics' Report: http://www.mckinsey.com/business-functions/mckinsey-analytics/our-insights/the-age-of-analytics-competing-in-a-data-driven-world
McKinsey Global Institute's 'How Artificial Intelligence can deliver real value to companies' report: http://www.mckinsey.com/business-functions/mckinsey-analytics/our-insights/how-artificial-intelligence-can-deliver-real-value-to-companies
-> there is rarely an analytics article, blog or book that does not reference MGI's seminal Big Data report of 2011 so anticipate similar industry-wide reverberations from MGI's latest works
D. Everybody Lies by Seth Stephens-Davidson https://www.amazon.com/Everybody-Lies-Internet-About-Really/dp/0062390856
-> Seth is a Harvard economist who shows how much you can learn about the psyche of individuals across the world from internet browsing and search data - brace yourselves for some shockers
E. Deep Thinking by Gary Kasporov
https://www.amazon.com/Deep-Thinking-Machine-Intelligence-Creativity/dp/161039786X
-> Gary - the legendary Chess master who lost to the machine in 1997 - argues for a focus on 'Augmented Intelligence' to enhance humans in their work, vs. an obsession with Artificial Intelligence
Also, has a summary Ted talk: https://www.ted.com/talks/garry_kasparov_don_t_fear_intelligent_machines_work_with_them
For the tweeters, I have a twitter list of 'Analytics Thought Leaders' around the world which you can find under my handle: @avipatch
2) Different techniques and algorithmic approaches
A. Introduction to Statistical Learning by James, Witten, Hastie, Tibshirani
-> Needs some maths background (though light on linear algebra) but gives readable explanations of all major techniques, the limitations and what questions to ask with model results
B. Master Algorithm by Pedro Domingos: https://www.amazon.in/Master-Algorithm-Ultimate-Learning-Machine-ebook/dp/B0147SEZ92?_encoding=UTF8&btkr=1&ref_=dp-kindle-redirect
-> Pedro's thesis is bold: development of machine learning is really exploring all feasible methods through which learning happens i.e., we are gradually building the learning architecture of the human mind!
C. Georgia Tech's Machine Learning course on Udacity: https://www.udacity.com/course/intro-to-machine-learning--ud120
-Time commitment: 10 full days. Lectures are good conceptual explanations of the different techniques
D. Data science for Business by Provost and Fawcett
-Comprehensive overview of major techniques, explained in English (as opposed to lots of Maths). Great read for Translator or Analytics Product Manager role
3) Building your own predictive models
A. Sign-up to DataCamp and learn R and Python: https://www.datacamp.com/
-Time commitment: basic + intermediate courses in R take 10 hours. There's another eight R courses to specialise, each taking half a day. I'm a Python fan for its versatility, but R tends to be preferred among data scientists, though its losing ground recently
B. Microsoft and Coding Dojo's Python for Data Science course on EdX
-Time commitment: 7 full days. Lots of practice at using Python to code ML algorithms, prepare data and run visualisation (requires basic Python knowledge)
C. Sebastian Thrum's Machine Learning course on Udacity: https://www.udacity.com/course/intro-to-machine-learning--ud120
-Time commitment: 7 full days. Takes you through step-by-step how to run, and interpret machine learning algorithms (requires basic Python knowledge)
D. For an advanced understanding of the mathematics (linear algebra heavy!) and rationale of different algorithmic techniques, see Elements of Statistical Learning by Hastie and Tibshirani: http://statweb.stanford.edu/~tibs/ElemStatLearn/printings/ESLII_print10.pdf
E. The number-1 text book on Deep Learning that (which is driving the breakthroughs in Artificial Intelligence): Deep Learning by Ian Goodfellow, Yoshua Bengio, Aaron Courville: https://www.amazon.in/Deep-Learning-Adaptive-Computation-Machine-ebook/dp/B01MRVFGX4?_encoding=UTF8&btkr=1&portal-device-attributes=desktop&ref_=dp-kindle-redirect
Lastly, my 4 favourite blogs for data scientists: Dataconomy.com; Datafloq.com; Datasciencecentral.com, KDnuggets.com. I also have a twitter list of the best analytics blogs under my handle: @avipatch
