Learning Library

← Back to Library

Data Science vs. Data Analytics

Key Points

  • Data science is the broad umbrella that encompasses all activities related to extracting patterns, building models, and deploying AI, while data analytics is a specialized subset focused on querying, interpreting, and visualizing data.
  • A data scientist (the role in high demand) follows a seven‑step lifecycle—identify problem, mine data, clean data, explore data, engineer features, build predictive models, and visualize results—repeating iteratively.
  • Core competencies for data scientists include strong machine‑learning/AI knowledge, programming in Python or R, experience with big‑data platforms like Hadoop or Spark, and solid SQL/database skills.
  • Data analysts concentrate on understanding existing datasets, performing descriptive analyses, and creating visual reports, positioning the role as a more focused, less model‑heavy counterpart to the data scientist.

Full Transcript

# Data Science vs. Data Analytics **Source:** [https://www.youtube.com/watch?v=dcXqhMqhZUo](https://www.youtube.com/watch?v=dcXqhMqhZUo) **Duration:** 00:06:29 ## Summary - Data science is the broad umbrella that encompasses all activities related to extracting patterns, building models, and deploying AI, while data analytics is a specialized subset focused on querying, interpreting, and visualizing data. - A data scientist (the role in high demand) follows a seven‑step lifecycle—identify problem, mine data, clean data, explore data, engineer features, build predictive models, and visualize results—repeating iteratively. - Core competencies for data scientists include strong machine‑learning/AI knowledge, programming in Python or R, experience with big‑data platforms like Hadoop or Spark, and solid SQL/database skills. - Data analysts concentrate on understanding existing datasets, performing descriptive analyses, and creating visual reports, positioning the role as a more focused, less model‑heavy counterpart to the data scientist. ## Sections - [00:00:00](https://www.youtube.com/watch?v=dcXqhMqhZUo&t=0s) **Distinguishing Data Science from Analytics** - The passage explains how data science serves as the broad umbrella encompassing tasks like data mining and model deployment, while data analytics is a specialized, visualization‑focused subset within that broader field. - [00:05:53](https://www.youtube.com/watch?v=dcXqhMqhZUo&t=353s) **Data Science vs Data Analysis** - The speaker contrasts data science’s broad, algorithm‑heavy workflow—from data collection to custom predictive modeling—with data analysis’s narrower focus on answering specific questions, noting that mastering both ensures you can keep inventory items like cantaloupes in stock. ## Full Transcript
0:00Data science and data analytics. Are they the same thing? Well, you may have seen these terms 0:08used interchangeably, but if I'm mining data from  a large dataset, am I performing data science or 0:15am I performing data analytics? Or what if I'm  trying to create a prediction of when my store 0:20will sell out of our current inventory of  cantaloupes? Well, is that analytics or is 0:27that science? It's worthwhile to understand the  difference to better comprehend what these two 0:33fields can do, and also, if you're considering  a career in either field. After all, these are two 0:40different jobs. Somebody who works in the field  of data science is known as a data scientist. 0:49For data analytics, that role is called a data  analyst. Now, this is kind of a trick question 0:59because we can classify everything-- data  mining, data forecasting and all the rest of 1:05it --as simply data science. And that's because  data science is the overarching umbrella term 1:12that covers tasks related to finding patterns in  large datasets, training machine learning models, 1:17and deploying AI applications. Data analytics, it  could be argued, is one task that resides under 1:26the data science umbrella. It's a specialization  of data science, and it focuses on querying, 1:34interpreting and visualizing datasets. Data  science is iterative, meaning data scientists form 1:41hypotheses and experiments to see if a desired  outcome can be achieved using available data. 1:47And that is a process that is known as the data  science lifecycle, which usually follows seven 1:56phases. So first is to identify a problem or an  opportunity. Then the next phase is data mining, 2:06which is to extract data relevant to that problem  or opportunity from large datasets. Now that data 2:12will likely consist of a bunch of redundancies  and errors, which is fixed in the next stage, 2:18data cleaning. And then at that point, we move on  to data exploration analysis to try to make sense 2:26of that data. We'll then apply feature engineering  using domain knowledge to extract details from the 2:34data. And predictive modeling comes next to use  the data to predict or forecast future outcomes 2:42and behaviors. And then finally, we have data  visualization that represents the data points with 2:49graphical tools such as charts and animations.  And so the lifecycle repeats. Now, the role of a 2:54data scientist is an in-demand profession right  now. If that's something you're interested in, 2:59you'll want to develop deep skills in machine  learning and AI. It's helpful to be able to 3:04write code in languages such as Python--also in R--  it's another popular language for data science. And 3:14you should have experience working with big data  platforms. So perhaps Hadoop or Apache Spark. And 3:22it's also very helpful to have database knowledge  and SQL. So that's data science. But what about 3:30its specialization, data analytics? Well, the job  of a data analyst is to conceptualize a data set 3:39as it currently exists. So we have some data  here and we need to do something with it. And 3:48we need to be able to make decisions based on  this data. How do we conceptualize it? Well, 3:53four ways. One is through predictive analytics,  which helps to identify trends, correlations and 4:00causation within datasets like forecasting when  those cantaloupes would have all flown off the 4:06shelves. Or, in health care, to forecast regions  which will experience a rise in flu cases. There's 4:13prescriptive analytics, and that predicts likely  outcomes and makes decision recommendations like 4:19predicting when a tire will wear out and need  to be replaced. There's diagnostic analytics 4:26that helps pinpoint the reason an event occurred.  Manufacturers can analyze a failed component on an 4:31assembly line and figure out the reason behind its  failure. And then there is descriptive analytics 4:38which evaluates the qualities and quantities of  a data set. A content streaming provider might 4:44use descriptive analytics to understand how many  subscribers it's lost or how many it's gained over 4:49a given period of time and what content is being  watched. And while a data scientist is a clearly 4:56defined and specialized role, virtually any  stakeholder can be a data analyst. For example, 5:03business analysts can use BI dashboards to conduct  business analytics and visualize KPIs. But many 5:10organizations do employ professional, dedicated  data analysts, responsible for data wrangling 5:16and interpreting findings like why a company's  marketing campaign didn't meet expectations. If 5:22you want to be a data analyst, it helps to have  both analytical and programming skills. So this 5:29includes familiarity with databases. Also, you'll  need to know about statistical analysis. And also 5:38data visualization is another important skill.  So data analytics is often more focused on using 5:47statistical tools and techniques to interpret  existing data and offer actionable insights. 5:53It's usually less concerned with creating  new algorithms or models. Data science, 5:58on the other hand, has a broader scope that can  involve complex machine learning algorithms, often 6:03created from scratch. Data science focuses  on phases from data collection to predictive 6:09modeling. Data analysis, on the other hand, is  more about answering specific questions with 6:15that data. And if you've done both your  data science and data analytics right, 6:22you'll always be able to keep cantaloupes  and just about everything else in stock.