cool hit counter EverybodyLies_Intefrankly


On the Tim Ferriss podcast I loved to listen to for a while (yes, not anymore, he talks too much), the question he used to love to ask his guests was, what's your favorite book to give away (but not write yourself)? Since Tim is sort of active in the tech scene, I find it a bit amusing every time I hear that giving away books is still a common behavior in Silicon Valley?

This book was given to me by a very geeky friend of mine, and I remember thinking of Tim Ferriss the moment I got it.

The book's author, Seth SD, is an economist and a columnist for the New York Times.

The cover of the book is still very much at a glance, talking about big data and the human behavior that responds to it. Not so much related to economics, more sociological in nature. Since he is a columnist for the New York Times, he naturally writes with humor and common sense and knows how to catch people's attention with some of the most interesting topics to the general public.

Seth begins by explaining the very esoteric-sounding science of "data analysis" - using the example of how his grandmother described the best person for him to marry based on her life experiences to illustrate that "data analysis" is a discipline that humans have mastered and used in practice for years.

A really good approach to data science is often quite intuitive, because at its core it is nothing more than finding patterns in the data and anticipating from it how one variable affects another. People do this all the time in their lives.

(This book was delivered to my heart, and I think it was also an accurate judgment made by my geek friend who observed and accumulated some of my data ah)

(It is also important to emphasize that Seth also illustrates that while data science methods are often quite intuitive, the results are often the opposite of intuitive)

The database that Seth relies on primarily is Google's search, and while there are some other data sources interspersed throughout the text, Google accounts for about 80% or more of the content.

Cutting to the cover's conclusion that everyone lies, Seth uses a number of examples to illustrate, including.

Number of sexual acts (survey responses and actual condom sales)

Racial eradication (the volume of searches for statements that are clearly discriminatory, especially in relation to political events, such as the Obama election)

Election participation rate (survey responses and actual searches of polling places)

and other examples. He concluded.

Search data shows that the real world we live in is very different from the academic and media portrayals that rely on public surveys.

Seth sees big data as manifesting itself in four main forces.

Get brand new data

Have authenticity and honesty (the internet is the tree hole of the new world)

Have data on a new niche of people

Greater assurance of arbitrariness in experimentation

Some interesting data findings from Seth are at the end of the article.

The most interesting part of his book for me personally was chapter 3, advocating careful handling of data.

He offers some warnings, mainly from the perspective of data science itself, such as the latitude trap (too much latitude can sometimes mislead the interpretation of data).

Then there may be a gap between what the data measures and what you really want to know. Interestingly, Seth uses the example of Facebook, and while Facebook can understand data about clicks, "likes", retweets, and comments, the most important questions, such as how the whole experience was, whether it helped the user understand the world, helped the TA connect with friends, or even whether it made the TA laugh, etc., the data can't effectively answer.

Seth's advice is that big data needs small data and people to help.

Some of the data is not available, and in addition to the ethical issues themselves, Seth raises questions from the data side. Facebook, for example, found that likes correlate well with IQ, extraversion, and conscience. For example, people who like Mozart, thunderstorms and curly fries usually have a higher IQ, while those who like Harleys, "I Love Being a Mom" and a country band usually have a lower IQ, but can such data be used to determine hiring?

Seth used several examples of how this data can currently be used for "big data", such as increasing protection for Muslim neighborhoods if there is a rise in hostility towards Muslims, but not taking action against individuals who search for individuals who express hatred or even kill Muslims, because even looking at the data, only a maximum of 10% of people actually engage in hateful actions.

The entire book is written in the style of an easy read. At the end of the book Seth says he was first led down this path by The Devil's Economics. Seth's writeup could also be considered evasive on the issue of how to avoid misuse of data analysis by companies or institutions. In terms of how to use big data to create value in these instances, he uses interesting life examples, such as deciding whether an athlete has a better athletic career ahead of him, who to bet on in horse racing, how casinos decide when to intervene when gambling, etc. Some of the things that are already happening with big data to design urban transportation and combat the spread of epidemic diseases, all of which have a greater impact on society, are not mentioned at all.

You don't need a lot of data, you need data that fits.

Anti-depressants are not as effective as temperature. Expect to take medication, rather than move to a warmer place to be based.

Middle-class families living in the suburbs are more likely to produce NBA stars than low-income families living in the city, and their names tend to be more common. (In an interview when LeBron got his second NBA championship in 2013, he said I am Lebron James, from Akron, Ohio. From the inner city. I am not even supposed to be here. There was a lot of criticism on social media at the time, mainly because it felt a bit hypocritical for him to identify himself as a dark horse. (But this argument is actually consistent with data analysis)

Whether what is in Freud's Interpretation of Dreams, especially dreams shaped like sexual organs, has sexual connotations is not confirmed in the data analysis, especially, for example, dreams of bananas, cucumbers, etc., which are traditionally applied to Freud's theories and are consistent in the data with their own status in the fruit and vegetable session.

Another famous Freudian theory is about the association between slips of the tongue and the subconscious, and Seth's main analysis is penumbra. He shows, through a very informative analysis, that no differences were found with other penumbras.

Freud's "Oedipus complex" (Oedipus complex) is evident in the data analysis. His point about the lifelong impact of the childhood phase of sexual initiation is similarly confirmed.

How do you find out if the other person is interested after the first date? Women usually have more variation in tone, are gentler, and use less indifferent words like "maybe" and "probably". More importantly, women may talk more about themselves.

Where are the men? Men usually speak in a low, smooth tone (more masculine). Yet the data also proves everyone's intuition that whether or not a conversation is interesting plays a small part in the matter of whether or not a man is interested in a woman, with appearance being a much higher determinant.

Many of the important decisions we make in life, such as which team we like, our political beliefs, etc., are formed at a certain age and have a lot to do with the social and political environment of the time.

Getting into a good school is important, but whether or not you get into the best school is not as important. Compare the two groups of people who got into the best school with the lowest scores to the two groups who came close to getting into that school without a significant turnaround in their fortunes.

Read the entire book Thinking: fast and slow only 7%!

So Seth's last chapter was random as hell, meaning, I guess not many people saw it here either.

3、5 mustread introductory Python books have you read them all with benefits
4、Read the difference between data analysis and data mining in one chart

    已推荐到看一看 和朋友分享想法
    最多200字,当前共 发送