Data Analytics vs Artificial Intelligence

Happy Chinese New Year! After a week of Chinese New Year holiday, it is time to back to work!

Before we compare data analytics against artificial intelligence, we should have a quick look of their definition first.

Data analytics are the processes to examine raw data with the purpose of drawing conclusions about information they contain. It involves in the application of various statistical models or algorithms to derive insights to support decision making. This technique is widely used in commercial industries to enable organizations to make more accurate business decisions with strong evidence or support. The scope of data analytics can be broad, from basic statistical models, business intelligence (BI), online analytical processing (OLAP) to various advanced analytics. Here is an example of data analytics application:

•  Healthcare: The main challenge for hospitals is how to treat as many patients as they can in an efficient way (e.g. lower cost, shorter time, better treatment). Huge amount of data generated in hospitals is being used to keep track of the process and optimize patient flow, treatment and equipment use. The estimation is there will be a 1% efficiency gain that could produce more than $63 billion in the global healthcare savings.

(Ref: http://techarcis.com/big-data-analytics-for-healthcare-how-big-data-will-change-healthcare/)

For artificial intelligence, from a textbook definition, it is a sub-field of computer science, and its goal is to enable the development of computers that are able to do things normally done by human beings, and specifically, things associated with people acting intelligently. There are lots of definitions in the world, for example, some of them believe the goal is to build systems that think exactly the same way that people do; some of them do not really care if there are any human thought inside the computation, they just want to get to job done; while some definitions fall in-between. One of the famous examples is IBM Watson.

•  IBM Watson: Watson is an IBM supercomputer that combines artificial intelligence and sophisticated analytical software for optimal performance as a “question answering” machine. The supercomputer is named for IBM’s founder, Thomas J. Watson. Cognitive computing technology of Watson can be endlessly extended to various applications. The device can perform text mining and complex analytics on huge amount of unstructured data, it can also support a search engine or an expert system with capabilities far superior to any previously existing. One of its applications is to seek out new treatments for cancer patients.

(Ref: http://www.ibm.com/madewithibm/au/en/watson/)

Indeed, such two levels (data analytics and artificial intelligence) may not be detailed enough to categorize various analytics technologies. To drill down to a deeper level, there are four types of analytics:

• Descriptive Analytics (What happened?): To reveal what is happening now based on input data. Real-time dashboard or reporting tool are commonly used.
• Diagnostic Analytics (Why did it happen?): To look at the performance in the past to determine what happened and why it happened. The analysis result is often presented with an analytic dashboard.
• Predictive Analytics (What could happen?): To analyze what scenarios are likely to be happened. The deliverables are usually a predictive forecast.
• Prescriptive Analytics (What should we do?): To reveal what actions should be taken. This kind of analysis is the most valuable. Results include rules and recommendations for next steps (decision making).

(Ref: https://www.trendminer.com/self-service-analytics-subject-matter-expert/)

Descriptive analytics are the bottom tier of the analytic spectrum, but it does not mean that it is not important. It can be valuable to uncover patterns that offer insight. A simple example of descriptive analytics would be credit risk assessment. Analysis result of historical financial performance can be used as a reference for projecting the customer’s future financial performance. Descriptive analytics can also be used in sales cycle, for example, to categorize customers by their likely product preferences and sales cycle.

Diagnostic analytics are used to determine why something happened. For example, in a social media marketing campaign, descriptive analytics can be used to assess the number of posts, mentions, follows, fans, page views etc. There can be lots of mentions in various social media platforms while they can be summarized into a single view to see what worked in those previous campaigns and what did not, with diagnostic analytics.

Predictive analytics ride on the dataset to identify past patterns to predict the future. For example, some companies are using predictive analytics for future sales lead; some companies even go one step further to use predictive analytics for the entire sales process, analyzing all customer relationship records including number of communications, types of communications, social media linkage and other relevant documents. Having a good use of predictive analytics can absolutely help to support sales and marketing, and other kinds of complex forecasts.

Prescriptive analytics is very valuable, but it is not commonly used. According to Gartner [1], 13 percent of organizations are using predictive but only 3 percent are using prescriptive analytics. In healthcare industry for example, patient population can be better managed with prescriptive analytics by revealing the actual utilization and resources abuse, some measurements like adding factors filter (e.g. diabetes level) to determine where to focus treatment. The same prescriptive model can also be applied to most of the target groups or workgroup in almost any industries.

If we talk to data scientists, they would tell us how they first acquire data and how they have it cleaned, how will they transformed those data into a useful form and then using their knowledge to decide what kind of method or analytics technique will best fit for the problem they are tackling. All analytics mentioned above required a certain level of skill to extract or reveal useful knowledge from those data.

Being able to harness data analytics can deliver big value to business, adding context to data also helps to tell a more complete story. By reducing complex datasets to actionable intelligence, more accurate business decisions can be concluded. If you know how to demystify such big data for your customers, your value will then go up significantly!

[1] http://www.enterpriseappstoday.com/business-intelligence/gartner-taps-predictive-analytics-as-next-big-business-intelligence-trend.html

Original Link: http://datasciguru.com/2017/02/06/data-analytics-vs-artificial-intelligence/