Data Science: What is expected in 2019
Data science changes rapidly. New advances in AI and machine learning mean that data can be applied in a completely new way, and in an unprecedented system of modelling, to do far more than was possible a few years ago. Cloud also ushered in a new era of data science by making software more portable and versatile.
The Techopedia asks experts what we might see in the coming year. Here are some things that are likely to occur in 2019.
“The demand for smart analytics applications will redefine the practice of corporate data: Companies compete to become data-backed businesses, but only a small part of the sophisticated analytical value that has been opened. By 2019, there will be high demand for new innovations around intelligent analytics applications that are driven by real-time interactions, embedded analytics and AI. …
“The emergence of data engineers brought AI to the forefront of the company: Last year was the year of data scientists. The company is very focused on recruiting and empowering data scientists to create sophisticated analytics and machine learning models. 2019 is the year of data engineer. Data engineers … specialize in translating the work of data scientists into hard-driven software solutions for businesses. This involves making in-depth AI development, testing, develops and auditing processes that enable companies to combine AI and data pipelines on a scale throughout the company.
“Human and machine learning forms a symbiotic relationship to drive real-time business decisions: In 2019, the world of AI and analytics needs to meet to encourage more meaningful business decisions. This will require a general approach to combining historical batch analytics, streaming analytics, location intelligence, graph analysis, and artificial intelligence in a single platform for complex analysis. The end result is a new model for combining ad-hoc analysis and machine learning to provide better insights faster than before. ”
“Developers will learn that they need data scientist friends.
“Developers will not be data scientists – one writes code, one thinks in mathematics and models. But devs will increasingly need to understand data science methodologies and to integrate data science models into their workflows. Data makes software smarter, provides capabilities to predict results or anticipate user needs through machine learning. Developers can expose data scientist models through the API, and embed them in domain-specific applications to really drive change.
“Consider retailers who are trying to smartly decide where brick and mortar stores are to fulfil e-commerce orders. A data scientist can make a model that calculates the optimal store from which to send, so the company sends a sweater that is likely to sit on a shelf in a store in a warm location, rather than what a shopper might buy in a very cold area. The developer can pull such intelligence into the fulfilment application, and put it in the hands of employees to make the right decision. ”
– Siddhartha Agarwal, Vice President of Oracle Cloud Platform
“By 2019, artificial intelligence (AI) and machine learning (ML) will almost reach its full potential by connecting and processing data faster through the global distribution of edge computing platforms. AI and ML insights are always available but may make use of a little slower than needed on traditional cloud platforms or data centers. Now we can move computing and storage capabilities closer to where data is taken and processed, enabling companies, organizations, and government agencies to make wiser and faster decisions. We have seen this in the way airlines and aircraft services, government defence agents respond to hackers and how personal assistants make recommendations for future online purchases. This year, thanks to AI and ML, someone will finally know whether that special person really wants a fruitcake or an electric washing machine. ”
– Alan Conboy, CTO office, Scale Computing
“2019 as if it would be the year of analytics, machine learning and AI. These tools are available, although their extraction is often delayed due to failure to match this new capability with the appropriate new workflows and SOC practices. Next year should look at some fraudsters – those who claim to use this technique, but actually use the latest generation of correlation and alert – deviant techniques, allowing real innovators in this field to start dominating. This is likely to lead to several acquisitions, because large incumbents, who have struggled to develop this technology, are trying to buy it. 2019 is the year to invest in machine learning security startups that show real capabilities. ”
– Stephen Gailey, solution architect, Exabeam
“When AI and ML become mainstream, a new generation of security data scientists will emerge in 2019: AI and ML techniques depend on data. Preparing, processing, and interpreting data require data scientists to become a polymath. They need to know computer science, data science, and most importantly, must have domain expertise to be able to tell bad data from good data and poor results from good results. What we have begun to see is the need for security experts who understand data science and computer science to be able to first understand the security data available to us today. After this data is prepared, processed, and interpreted, the data can then be used by AI and ML techniques to automate security in real time. ”
– Setu Kulkarni, WhiteHat vice president
“In software development, the big story in 2019 is machine learning and AI. In the coming year, the quality of software will be as much as what the machine learns and AI can solve it as something else. In the past, the shipping process was designed to be lean and reduce or eliminate waste, but for me, it was an outdated, half-glass empty way to see the process. This year, if we want to make full use of these two technologies, we need to understand that the opposite of waste is valued and take a full half-glass view that being more efficient means increasing value, rather than reducing waste.
“After that perspective is embedded in M.O. we, we will be able to direct our views to be better through continuous improvement, be faster to react and anticipate customer needs. However, as we further integrate and utilize machine learning and AI, we will realize that increasing value requires predictive analysis. Predictive analytics allows simulating shipping paths based on available parameters and options, so you don’t have to ‘thrash’ organizations to find a path to improvement. You will be able to upgrade virtually, learn lessons through simulations and, if ready, implement a new release that you are sure will succeed.