What is Data Science?

In simple terms, data science analyzes information both visually and statistically to discover patterns and predict outcomes. Data science utilizes computer programming to convert large quantities of data into compressed visualizations in the form of tables, graphs, and more.

Information is the basis of the future. With millions of people using the internet every day, information is constantly being shared. And the first step to solving any problem is understanding all the information given. That's where data science comes into play. By converting large volumes of data into a more comprehensible form, we can make decisions based on facts, statistical numbers, and trends.

Data science is used pretty much everywhere. Most major companies use some form of data science to analyze information, which in turn helps them cater to their customers. But to understand the versatility of data science, here are some real world implications of the field.

Social Media

Every company or brand that has a social media account uses it to gather information about their customer demographic. By computing thousands of direct customer responses, companies can make informed decisions based on statistical inferences

COVID-19

Another huge implication of data science is in a pandemic we are all familiar with: COVID-19. Scientists use data science to find details about the virus such as where it has the biggest impact and possible vaccines.

There are 3 main parts to learning data science and each part plays a vital role in understanding the aspects of how to analyze and interpret data.

Computational Thinking

Serving as the foundation for data science, computational thinking includes using programming to perform simple tasks on data. For example, this can be putting information into a table or performing simple mathematics on selective parts of data. The focus of this step is to get a basic understanding of the data and formulating a hypothesis.

Inferential Thinking

Inferential thinking is all about statistical analysis using mathematics. Here, we test hypothesis by simulating the data that we previously organized. Now that we have a general understanding of the topic, we can run simulations on random samples of the data to get an even more in depth understanding of the numbers.

Machine Learning & Predictions

Machine learning is a branch of artificial intelligence that is based on the idea that machines should be able to learn and adapt through experience. By using algorithms to train models, we can predict the outcomes of specific sets of data, without defining a set rule or program. This transition from programming to utilizing AI redirects the heavy lifting off of the human and into a system that continuously grows smarter and faster.

To dive deeper into data science, we can use artificial intelligence and machine learning to further our understanding of a topic. In AI/ML, a model learns a program by analyzing huge samples of input and output data. Given these sets of data, the machine tries to understand the underlying program that converts the input into the output. Traditional programming, on the other hand, gives the computer the input data and the actions to perform on the data to change it into the output data. But, there are a multitude of ways to train models. And all of them, involve algorithms.

An algorithm is a specific way a model adjusts itself to perform better on new data. There are tons of different algorithms that can be used to teach models, but there are two main types of them: unsupervised and supervised. Unsupervised algorithms analyze and cluster unlabeled data sets without the need for human intervention. Supervised learning, however, requires humans to manually label datasets for the algorithm to function.

Popular AI/ML Algorithms

Linear Regression is a model that assumes that there is a linear relationship between the input variables and the output variables. You can think of it as plotting the inputs and outputs as ordered pairs on a coordinate plane, and then drawing a straight line of best fit through all of them.

A Decision Tree is used to solve regression and classification problems by predicting the class or value of the target variable. This is done through learning simple decision rules inferred from prior data.

The objective of K-Means is to group similar data points together and discover underlying patterns. To achieve this objective, the algorithm looks for a fixed number (k) of clusters in a dataset.