Data For Earth

Tag: ai

  • Big, Big Data

    In the past few years, data science has become more and more of a prominent topic in today’s society. It is ultimately what the foundation of this blog is built upon. However, people sometimes forget just what data is and what it could be. 

    The Merriam-Webster dictionary defines data as “information in digital form that can be transmitted or processed”. It is, of course, not a very broad definition but an accurate one nonetheless. 

    This then begs the question, what do we define as information? 

    Well, the answers are almost endless. Information and thus data can be crafted from almost anything you can imagine. Look around your room or wherever else you are and you’ll see just how much data you can find. The shape and dimensions of your room that make up the floor plan are data, the hexadecimal colors of your walls are data, the temperature, amount of clothes, number of books, and even how long your light has been on are all data.

    Because of the limitless possibilities of what we can define data as, there are almost limitless opportunities of what can be done with it. As a result, data science is used not only everywhere, but also for essentially anything. Being able to extract data from a process, a phenomenon, or a company shows us patterns and their subsequent implications on how to maximize efficiency. These days, anyone can use data for virtually anything. For example, corporate enterprises use data to advise investments and realtors use data to help set prices to sell a home. In this way, the possibilities for data science are endless and left up to the human imagination.

  • AI in Data Science

    AI in Data Science

    A lot of people in today’s world tend to confuse the concepts of “Artificial Intelligence” and “Data Science”, treating the former as a buzzword for the latter. However, while they are related, they are often used for different purposes. Data science is fundamentally the use of statistical tools and analysis to give meaning to a large set of data. AI itself may then use the patterns that are found in the data to create machines capable of performing tasks that would require cognitive input to do.

    While the concept of data science has existed for a very long time, it is the use of AI to enhance and implement what is derived from data which has seen immense improvement over the past few years. All neural networks and machine learning algorithms built require a training dataset. This dataset is broken down by the network until patterns are found which could be used to relate certain traits of the data to an expected output. For example, in a regression model for housing prices as a function of various factors (location, square footage, amenities, view, surrounding area, etc), patterns would be found using data analysis to teach the model how to analyze. The weightage of certain factors can be calculated using statistical analysis, but the idea that these patterns could be fed into a program to automate a task such as predicting housing prices is very much artificial intelligence. 

    Although this was a very simplified example, this is fundamentally what many of the largest companies in the world do in regards to creating models. For example, ChatGPT is a large-scale language model that uses a prompt to predict what order of words to generate to address said prompt. Examples such as these showcase how AI amplifies the capabilities of data science, making it more efficient and impactful across various domains.