The Reasons Why Data Science Is Indispensable in The Economics
How are data science, big data, and machine learning affecting the field of economics? Originally appeared on Quora: the place to gain and share knowledge, empowering people to learn from others and better understand the world.
Nowadays all of the technological fields you mentioned are indispensable in the economics and especially in the stock market and speed trading. Big data provides a large pool of information, which can be refined with data mining and then passed to machine learning algorithms, which are superior when it comes to analysis of complex unstructured data and fast decision making (deciding whether to buy/sell stocks by analyzing tweets for instance).
Here’s how trading was done in the past:
A bunch of people, sitting round the table reading various reports, news and documents in order to decide whether a particular stock is promising, then making wire calls to buy/sell them. As you can probably guess this was a tedious process that required lots of time, which in a field such as stock exchange was very limited.
These days trading has changed quite a bit.For simplicity we can think of the stock market as a huge chat room where people are exchanging opinions. This information can form trends and sometimes people aquire or sell stocks merely out of rumours that a certain company will increase/decrease in value (imagine someone quite respectable creating a trend for iPhone purchase by posting a very positive review on the chat, increasing Apple’s profits, boosting investor’s confidence and raising its share price).
This is where Big Data, Data Mining and Machine Learning comes into play
Let’s consider this real-life example:
- Twitter is a huge source of useful data for investors. There are millions of tweets everyday sharing opinions about products, companies or even countries – all of which can play an integral part when deciding a company’s value (let’s consider this to be our big data).
- There are tens if not hundreds of startups trying to analyse terabytes of that real time twitter feed in order to predict how a particular stock will change in value (let’s consider this to be our data mining).
- Finally since there’s a lot of competition, decisions have to be made fast. Much faster than any human could do with a wire phone or even a click of a mouse. Nowadays billions of dollars in trading are left in the hands of sophisticated trading algorithms, whom in order to keep up with the competition must continually evolve and Machine Learning is often used for that purpose.
Contributed by Saul Venskutonis, studied Computer Science at Newcastle University