3 Data Science Books That Will Inspire You
It can be very inspiring and fulfilling to work with data science or analytics, but as with any other job, it is almost impossible to be 100 percent motivated every day.
No matter if you just have a few motivational dips sometimes, or if you have completely forgotten why you got into data science in the first place, I hope you will find new inspiration and motivation from these three books. They all highlight data science and its possibilities from different angles, but common for them all is that they convey a sense of curiosity and joy of discovery.
Everybody lies : what the Internet can tell us about who we really are
Seth Stephens-Davidowitz has written a book about what people write into Google’s search field and how that data can be aggregated and analyzed. Some people might censor what they tell their friends and family, but few censor what they in their privacy ask Google.
The book takes up topics ranging from politics to sexual preferences and how this data can be utilized. Could it for example be possible to prevent homicides by tracking searches for “how to murder someone”? What about distributing resources for youth centers based on regional searches for teenage pregnancies and school bullying? Or just correcting some societal prejudices, i.e., that men have a higher sexual drive than women? According to Google, there are twice as many searches made about boyfriends that won’t have sex than girlfriend that won’t have sex.
The book shows what a gem data is for us to understand the world and ourselves better. I find it very inspirational to think about all the possibilities ways this information can be (responsibly) utilized, especially on a societal level.
Outliers : the story of success
This book by Malcolm Gladwell is not primarily written for a data science audience but the theme is very relevant for us, as it investigates outliers in the society. Why do some people become so successful at what they do? The book questions the widespread idea that success is merely a question of hard work.
Gladwell challenges us not to dismiss outliers, just because they are — outliers, but to instead invest more time into investigating them and to look further than to their most obvious characteristics. Gladwell brings up successful scientists and the importance of where and when they were born and what societal changes they were exposed to at a critical age and how they could benefit from this. Would Bill Gates have founded Microsoft if he would not have been young at a time when personal computers saw the light of the day? What about if his school would not have had a computer for the children to use (which was very uncommon at that time)? Or if he would not have been able to cheat the system at University of Washington to get unlimited access to their PC?
Gladwell gives several similar examples in the book, i.e. a theory on why many Asians are good at math and why The Beatles gained such a huge popularity.
I think the book is very inspiring, as it challenges us not to dismiss outliers and peculiar patterns but to set aside time and dig deeper. It is possible that outliers can give us insights that are not only connected to them, but can be useful in a much broader sense.
Data Smart: Using Data Science to Transform Information into Insight
This book is not about what data can tell us or how to better turn data into information, but how data science can be done in an alternative and more intuitive way.
The writer, John W. Foreman, has worked as a Data Scientist at the marketing automation platform MailChimp and wants to debunk the myth that machine learning is something exceedingly difficult that only math pros can do with the use of complex tools.
The book presents several common ML algorithms (naive Bayes, linear regression, k-means clustering…) and how they can be used to solve different business problems in Excel. Yes, you heard right, in Excel! The idea of the book is, however, not to tell us to start doing machine learning in Excel, but to show how easy it is to follow the logic of a model if you do the steps one by one in a spreadsheet.
This is a great book for those who haven’t calculated all the different models manually, those who want a recap, those who want to explain ML models to business users, and of course, for business users themselves.
In the midst of discussions about optimal ML models, feature selection methods, parameter tuning and new cool frameworks, I think it feels refreshing and inspiring to see how much can be done with just the basics and a focus on clarity and “good enough”. This in contrast to aiming for maximal accuracy/precision/whatever your measure is to the price of maximal complexity.
