When we talk about Artificial Intelligence and Data Science, it is incomplete if we don't add Machine Learning and Deep Learning in it. That's because they are interrelated and dependent on each other. So that you better understand what the differences and correlations between these four things are, consider the following brief explanation:
ARTIFICIAL INTELLIGENCE (AI)
The scope and implementation of AI is very broad, it even tends to be biased because several sources have their own opinions. However, the notion of AI can be summarized into 2 main concepts, the first AI is something that computers cannot do now but will be able to do someday (because they are always learning). The second definition of AI is intelligence demonstrated by machines. Intelligence / intelligence is the ability to acquire and apply knowledge and skills .
Humans develop machines to gain knowledge efficiently because machines can search/process information faster than humans (very basic example: calculator). However, machines cannot do things on an initiative. Humans have to give 'what to do and how to do it' to machines to work. The bigger hope is that humans provide a little 'knowledge', then machines can learn from that knowledge and develop themselves. True AI needs to be able to learn . So how do machines usually 'learn'? we usually know this with the concept of Machine Learning.
MACHINE LEARNING (ML)
As can be seen in the Venn diagram above, machine learning is part of artificial intelligence . Machine learning in short is a sub-field of artificial intelligence that deals with the design and development of algorithms. Machine learning can also be defined as a technique that enables improved performance on multiple tasks through experience. The focus of machine learning is to gain insight so that it can make data- driven decisions ( machine learning uses data to answer questions). For some machine learning casescan become so complex that additional methods are needed so that machines can imitate the workings of the human brain, also known as deep learning.
DEEP LEARNING
Although it sounds more 'wow' compared to machine learning , deep learning is not a weapon to solve all data-driven problems. Deep learning will not replace all machine learning algorithms or other data science techniques . However, deep learning can indeed solve more complex problems such as computer vision (the ability of machines to recognize objects in image data), speech recognition (recognizing voice data), and natural language processing (recognizing text data) using artificial neural networks (ANN). Simply,deep learning mimic the way the human brain works through a network of neurons (neural network) whose architecture is very diverse.
Then, what do the three terms above have to do with data analytics, data science, and big data ?
Big data itself is just a data processing concept whose size is very large including 7V: volume, velocity, variety, variability, veracity, visualization, and value . Big data is just a thinking concept and its understanding is also quite biased like artificial intelligence . However, it can be said that people who solve complex problems using very large data (meets 7V) through computer processes or techniques, have used the concept of big data.
DATA SCIENCE
On the other hand, data science has a different but related meaning to the above terms. Basically, everything a Data Scientist does includes the following three branches of knowledge:
- Computer science/programming , namely the ability to write, test, repair, and maintain code on computers. Programming languages that are often used by a Data Scientist at this time are Python and R (analysis) and SQL (accessing data).
- Mathematics/statistics , is the science that is the basic foundation of how an algorithm works. Some of the branches of science that are most often used are calculus, linear algebra, and probabilistic.
- Business knowledge , adapted to the background of the data and problems analyzed. Business knowledge is needed by a Data Scientist to make the right decisions based on the information obtained from the analysis.
Talking about "which is most needed today", of course all of these things are needed. However, because the domains/scopes are different, it is quite difficult to compare the terms above as a whole. In the era of big data like now, data is very abundant and the industry really needs people who have the ability to extract ' insight ' from that big data. In addition to seeking insight from the data, it should also be summarized in such a way that the output can be well informed for decision making in business. The work is done by a Data Scientist. According to a Harvard business review, Data Scientist is the most “ sexy ” job of the 21st century.
To learn and explore the work of a Data Scientist you can do it independently or by taking a course. One of the face-to-face data science courses is the Data Science School Algorithm which is the first data science school in Indonesia to have RStudio certified teachers. Algorithm provides a place to learn data science for anyone, not limited by any educational or scientific background. At the end of approximately 3 months of learning, students who graduate from Algorithm also have the opportunity to take part in Data Career Day which aims to help graduates find jobs in the realm of data science.