Big data is like teenage sex: everyone talks about it, nobody really knows how to do it, everyone thinks everyone else is doing it, so everyone claims they are doing it...
So we will not do talk about what the big data is, we will talk about how to use it.
Consider you are self employed and you are the owner of a small company with only two person working, you and your wife. Do you need big data analysis? The answer is no. It is like a driving bicycle, do you need any sensors in this case?- definetly no. So if you are the bycicle driver you can close this page just now and don't waste your time. You are continue your reading?
So you are the manager of midsize or large company.
What does it mean to manage mid size or large company? Managing a midsize or large company is like to drive a car. Would you drive a car without any sensors? - certainly no. And why? Because the sensors is your connection to external world, checking sensors you correct your driving. And if you driving a car without any checking with sensors - the wreck is inevitable. So in managing a company, if you will try to manage your company without any analysis of your data, coming from all sources, your records, your customers, market etc., the bankruptcy is inevitable. And where is the road sensors, you ask? Ok, sensors in your car is the same as indicators of your company. And indicators? Indicators is a result of your data, big data, analysis. But why the death of Titanic?
Because of the big data, or rather because of their lack of. The ship had a lot of sensors allows to analyze a lot of data a lot, but not all, and as a result - collision with an iceberg. How to use it?
Remember the fate of the Titanic , collect and analyze all available data! The absence of the most insignificant at a first glance, information could make a crash to your company. Only the collection and analysis of all available information will help to avoid a "collision with an iceberg". And how does this big data analysis works?
Very simple, in the same way as sensors in your car, just transforming the information from external world into understandable indices. It's look like more questions than answers.
But do you really interesting how works the sensors in your car, probably no. All you need to know in this case that your sensors is:
- works based on some physical and mechanical principles;
- works correct.
We will not waste the time discussing about physical principles in sensing the data in your car sensors, but will take a little piece of time to discuss about principles and methods which is lying in the base of data, big data analysis. Topological Data Analysis.
Because data has shape and the shape matters. Machine Learning.
The human mind cannot explore and analyse a high dimensional data (all data you need to analyse is high dimensional), so we have to use computer's capabilities, i. e. machine learning. Mathematical Statistic.
Machine learning works on bayesian, markov and neural networks, i. e. mathematical statistic.