Week 4 - Paradigms

Paradigms

What is a paradigm?

To start of this blog, we first need to understand what a paradigm is, then how it relates to science.

owlcation.com has a fantastic description of what a paradigm is:

 "Essentially, a paradigm is a set of assumptions governing how we interact and interpret the world."

Basically a paradigm is the assumptions which we generally believe to be true, it's a perspective of what we think is real and true that we base our thinking on. Everyone has their own paradigm which reflects their beliefs and experiences.

In relation to science, a paradigm is a shared set of assumptions and beliefs. Scientists can't explain the universe and how it works, so they form a paradigm to base their thinking on. Without this base you can't start because there is no starting point.

Paradigms don't always stay the same. Sometimes the assumptions can be proven wrong, and new scientific progress is made. New sets of assumptions are made. This change is called a paradigm shift.

Big Data

The paradigm I chose to talk about is big data. Big data focuses on the analysis of massive datasets that cannot be processed with traditional systems. Big data can be structured, semistructured, or completely unstructured.

Traditionally scientists had to manually process information and it took huge amounts of time to perform calculations and test formulas and ideas. With the rise of computers many things could be simulated with computers, which sped up progress massively, but scientists still had to make sense of the data themselves.
Big data goes to the next level and uses machine learning/AI to process masses of data and find patterns. With big data we collect data from many different sources in bulk, and let the computers process it rapidly to make predictions, understand behaviors, and more.

What can it be used for?

Big data has a number of applications. At its core, big data is just analysing huge amounts of data. Specifically, this usually means finding patterns. For example, doctors can collect data of individual patients over time.
The doctors can't look at every patient's history and predict what will happen, but by putting all the complex data into a computer, the computer can compare patients, and find patterns in each individual patient's medical history.
Based of this, after gathering enough data the computers can start making new data in the form of predictions, something that doctors can never do accurately.
The more data the computer accumulates, the more accurate the predictions get.

Big data can make computer analytics more organic. Rather than coders having to make algorithms to analyse specific things, they make good machine learning algorithms which learn how to process the data themselves.

Big data is also great for visualising and simplifying hugely complex datasets for humans to understand.

Why do I like it?

The reason I quite like big data is because it depends more on technology and data than human reasoning. Humans can be biased and make mistakes, and they can only comprehend and process so much. Technology constantly improves and makes less mistakes than humans, and massive supercomputers can find patterns in data than humans might never find. By using big data we can solve many problems and make predictions that would otherwise be impossible.