A question that comes up often in conferences and events that theDevMasters has attended is: what’s the difference between data science, machine learning, and artificial intelligence?
The best way to answer this question is coming from a perspective of who usually gets tied into these topics the most: data scientists.
To start, data science is the process of extracting information from data through tools. A data scientist is expected to understand data science: collecting data, cleaning data, running analysis on data, and using various tools to answer or perform the project needs. One of those tools is machine learning. Not all data scientist utilizes modeling to solve business situations, therefore data science only encompasses a small portion of machine learning. Additionally, while artificial intelligence is a section of machine learning, not all data scientists create artificial intelligence driven results or end products, nor fully work with artificial intelligence tools.
To broaden further, machine learning is considered the process to further automate processing of decision making through statistical driven but computer programmed algorithms. Algorithms can range from supervised learning, where the results are already known or need to be recreated, or unsupervised learning, where results are not clear and the data needs to be investigated for further patterns or scenarios. Data scientists are welcome and lately, sought after due to their ability to understand machine learning more than their data analyst counterparts.
Within these two subsets of machine learning is where artificial intelligence continues forward. Artificial intelligence narrows deeper into what machine learning can do, but without the borderlines of supervised or unsupervised. Therefore, sometimes artificial intelligence is synonymous with deep learning. Moreover, artificial intelligence brings semi-supervised learning and reinforcement learning, which can distinguish between processes for further analysis than typical machine learning.
Artificial intelligence, in this stage, is not to be confused with artificial general intelligence, which is the next stage in data and robotic products: machines having the capacity to be creative. All artificial intelligence now is reliant on predefined data and experiences, and is not expected to create new scenarios with limited data, nor create new data that is outside of the scope of what it has been able to analysis to then create such new scenarios. In the future, it will and data scientists will be the driving factor behind this radical change.
Overall, to bring it back to the present, data scientists will perform data activities that a data analyst does, with further alignment with what the business goals are necessary with data science, and utilize tools like machine learning or artificial intelligence to achieve those goals.