Key data science trends 2017

João Pela, a data scientist at RotaGeek and former CERN employee sees the growth of data by sensors as a huge opportunity for businesses. He says: “Internet of Things is becoming smarter, and so smart objects are making their way into the mainstream consumer space.

"Machine learning is key to unlocking the full potential of this field: in the future, not only will we all have a personal assistant to help with everyday situations, the smart objects will themselves know how and when we like to use them (Google Nest, for example, is already making way in this field)."

Liz Curry, business process manager at Comic Relief agrees: "The amount of data that’s available because of the Internet of Things means that it’s possible to find more hidden insights than ever before."

One challenge for businesses is accessing the necessary skills to make the most of the large volumes of informations available to the. "We are seeing the increasing struggle to find the right people with the background, skills and real world experience for these roles," says Jessica Kirkpatrick, a data scientist at Hired. "With many seeking traditional backgrounds such as data science and statistics, companies are being forced to compete over a limited, highly skilled talent pool, especially at a senior level."

Read next: How to get a job as a data scientist: What qualifications and skills you need and what employers expect

Kirkpatrick advises that businesses get creative when it comes to hiring for data science roles. This means "looking outside traditional requirements and to alternative backgrounds. By seeking smart candidates in areas such as science, finance and business who have an aptitude in problem solving, businesses can give themselves access to a wider pool of untapped talent who have a foundation that they can build on."

Liz Curry, business process manager at Comic Relief believes that regulation will have a huge impact on the way businesses do data science this year. She says: "With the introduction of the GDPR in May 2018, I think the challenge to data analysis will be to ensure that we’re transparent about what we’re doing with our customers’ data whilst still trying to get actionable insights from them to benefit the sector."

Andrew Day, chief data officer at Sainsbury’s wants to see data science approaches reach more corners of the business.

"Where most organisations have been talking about the use of data science in solving customer and marketing-related problems, the step change will come in using maths ubiquitously across the organisation – in product development, pricing, ranging, site location, logistics and so on," he says.

"As a consequence, everybody involved in the management of a business will need to become familiar with the art of the possible and expect deploying mathematical and data-led solutions to become part of the day job."

Vince Darley, VP for growth at Deliveroo has a similar view for the year ahead, and sees a structural change on the horizon.

"This year I think the biggest shift will be in how broadly many businesses will begin to see the benefits of data science, and how that will be driven by more businesses changing from a single central data science team to a more productive, impactful structure where data scientists are embedded across the business in many, many different operational and product teams."

Read next: Best data science tools: Data science platforms for modelling and deploying machine learning and predictive algorithms

Gideon Mann, head of data science at Bloomberg sees image recognition as the boom area in data science communities this year. "There is so much technology developed right now which hasn't been deployed. I think over the next year the deployment of that existing technology will be huge and if I was to bet on an area of growth it would be image recognition."

Read next: How Bloomberg is using machine learning and data science to keep users hooked to its terminals

This follows the open sourcing of Google's TensorFlow machine learning software library, which was developed in house to help the search giant improve the way it categorises and organises image files.

Kjersten Moody, vice president, information and analytics at Unilever says: "Focus in 2017 will be on joining up structured and unstructured data to use in prescriptive analytics."

Read next: How to use data scientists and machine learning in the enterprise

In short, businesses like Unilever are looking to move beyond just running analytics on their data to giving business users actual, prescriptive insights to improve business outcomes.