A Quantitative Perspective Of The Uncertain World

Zheng Lei / Wen

We live in an uncertain world, which is a major challenge facing mankind all the time. As the world changes faster and faster, the uncertainty is obviously enhanced, and the data is becoming more and more, we have to improve our self-awareness and adaptability. These two books on "data based decision" provide us with effective and easy-to-use tools. Among them, "data based decision-making" has been translated into four languages and published. From the height of methodology, the author redefines quantitative analysis, and explains with a large number of examples how to adopt appropriate quantitative methods to form our common sense. Through these simple methods, uncertainty can be effectively reduced. "Data based decision 2.0" introduces the decision-making method based on the collection and analysis of network big data. Compared with the previous book, this book needs more professional analysis tools to get more accurate and detailed results.

The theme of these two books is the quantitative decision-making method in the era of big data, which belongs to decision science. The method given in the first book requires some creative thinking, but it's easy to use and can be done with only pen and paper. For example, the ancient Greeks could estimate the circumference of the earth by the distance between the two places, and Fermi roughly estimated the power produced by the first atomic bomb test with some scraps of paper. There are a lot of interesting tips that are very useful. You may think that difficult theoretical problems are easily solved by the author. For another example, the author mentioned a five person rule, which is the best example of rapid quantitative analysis through simple calculation of statistics. If the company's business needs more telecommuting, we need to consider the average amount of time each employee spends on communication each day. If we don't need accurate results, we don't need to conduct a survey on all employees. We can just select 5 employees randomly. Assuming 30 minutes, 60 minutes, 45 minutes, 80 minutes and 60 minutes, it can be concluded that 93.75% of the median time spent by all employees is between 30 and 80. You can make decisions based on that without worrying about big problems. Because very few employees will be far away from the median time. This method is simpler and time-saving than small sample sampling. If you want more accurate results, you can select more samples, but the accuracy of the results is not improved.

Many times in life, we need this kind of situation that we don't pursue too accurate judgment. Through simple calculation and analysis, we can further determine the direction of our work, and the data in our hands is usually enough for this kind of analysis. Nevertheless, whether to carry out quantitative analysis still needs to be considered, because whether it is to conceive a simple and ingenious quantitative method or to use big data collection and analysis software for analysis, there are cost and benefit considerations. The reason why we do this is to reduce the loss caused by uncertainty. The author gives five steps: first, define the problems and related uncertain factors that need to be decided, determine what information (data) you know now, calculate the value and cost of collecting other necessary information, and then choose to apply the relevant quantitative methods to the quantification with high added value, make decisions and take actions.

For big data analysis with higher requirements, there are now very good algorithm software, which can monitor and capture the insightful information left on the network at any time. Moreover, the information is public. Generally, there is no legal or ethical conflict such as crawling private information by web crawlers. At present, many organizations have improved their decision-making methods and gained competitive advantages through this quantitative analysis. For example, the New York Police Force's Facebook monitoring team can use the information in Facebook to track down and convict criminals who killed innocent teenagers in gang struggle without any witnesses. Hice Messenger, a communication software made in India, adopts precise social media analysis tools to feed back the analysis results to the product R & D department to develop new functions. In less than three years, the number of users surpassed Facebook and became the popular communication software after WhatsApp. In 2016, rongwen group accurately predicted the outcome of brexit referendum and Trump's victory in the general election by analyzing social media information. However, the traditional public opinion survey gets the opposite result, which shows that the network big data analysis can analyze this kind of complex situation more accurately.

We live in a world flooded with open data, but a large amount of data is still ignored, which is a huge "gold mine" that has not been fully paid attention to. The Internet has changed our way of life, but the decision-making methods people use are still in the past. It is necessary for us to combine the decision-making methodology provided by these two books to make a preliminary analysis of the problem through a simple, fast and low-cost way to clarify the key points and difficulties. If necessary, we should concentrate the computational resources and use the big data algorithm software to collect and analyze those key links or information in real time, so as to improve the accuracy and efficiency of decision-making.