20 Data Trends for 2020

20 Data Trends for 2020

Though we cannot tell what the future holds for us, we can make predictions based on trends. Read about the key data trends from twenty thought-leaders for 2020.

JT Kostman Ph.D – A global cyber crime pandemic ($6T annually) and an ever-expanding alphabet soup of data privacy/protection legislation will increasingly require Data Scientists to accept dual responsibilities as data fiduciaries. As the volume, velocity, variety, virality, and viciousness of cyber attacks inexorably increases, AI solutions will increasingly become the only way to compensate for the projected shortfall of 1.8 million cyber security professionals otherwise needed to combat increasingly sophisticated and determined adversaries – and keep corporate executives out of court.

Cassie Kozurkov – We can expect to see improvements in tools for data science as more user experience designers take an interest in the data scientist as user. Image data will grow in importance as the camera becomes more than a way to capture memories, but evolves towards a more natural way for users to interact with apps. My favorite example of this trend is product search and translation in Google Lens.

Ben Lorica– Reinforcement Learning (RL) begins to show up in enterprise applications. Tools for RL are becoming more accessible: case in point, RISELab’s Ray framework has an RL library (RLlib) that is popular among practitioners and developers. This is beginning to translate into enterprise use cases. Netflix, Google (Youtube), JD.com, and Facebook have described how they’ve incorporated reinforcement learning into production recommender systems. There are also early applications of RL in business process optimization and simulation, both promising areas for RL in the enterprise.

Kirk Borne – In 2020, we can expect intensified focus on three big trends: (1) Automated Machine Learning (AutoML) tools and training to empower diverse teams of citizen data scientists; (2) Renewed investments in data curation activities (data hunting, collection, cleaning, tagging, indexing, federation, fusion, and orchestration) to fuel enterprise-scale ML and AI training, deployment, and sustained operations; and (3) staff re-skilling and up-skilling efforts in the areas of Data and Digital Literacy at all levels within organizations.

Bernard Marr: Here are my top 3 AI / data science trends for 2020: I am seeing 2020 as a year where the fields of big data, IoT and AI will continue to fuse, and where optimized AI chips and 5G will enable amazing innovations. This will mean:AI as a service will enable any business to leverage machine learning.
Augmented analytics will automate more of what analysts and data scientists do today.
AI will move to the edge and into everyday devices, which in turn will give rise to the Intelligence of Things.

Jordan Morrow – The world of data and analytics is expanding, and as companies have taken notice of the need of data literacy, the trend in 2020 will continue into one key aspect of data literacy: data-informed decision making and decision science.  Being able to build a great data viz, analyze and read data is one thing, but it has to translate into smart decision making.  Data-informed decision making will be a key trend in 2020, and decision science where we combine the data and human element together to create success.

Matt Dancho – Moving into 2020, three things are clear – Organizations need Data Science, Cloud, and Apps. Software and Data Science are moving closer together. This is driven by business needs to democratize data-driven decision making. To keep up, we (Data Scientists) need to learn ML + Cloud + Apps. 

Andriy Burkov – On the “beating-the-benchmark” front, I expect BERT, the language modeling neural network model that increased the quality of NLP on virtually all tasks in 2019, to shine even more in 2020. This year, Google started to use it as one of the major signals of relevancy — the most significant update for many years. In technology, the key trends, in my opinion, will be even wider adoption of PyTorch in the industry, increased research on faster neural network training methods and fast training of neural networks on convenience hardware.

Andreas Kretz – In 2020 we are going to witness the failure of many data science projects as companies struggle to turn their proof concepts into real business.

Kristen Kehrer – In 2020, the new trend is going to be interpreting and communicating ML models to the business. There has been talk about the “data translator” role, rather than focusing on professional development for current DS/ML practitioners, but this is a real need. A large percentage of machine learning projects fail to become adopted, partly due to a lack of the business understanding these “black box” models.

Ben Taylor: 2020 will be the year autoML goes mainstream with competition heating up from companies like Google, Amazon, Data Robot, H2O and many others as they fight to make their offerings part of the standard toolkit. Another trend we will see is an increase in data set complexity where fusion models (structured + unstructured data) become more common. We will also see improvements with storytelling around deep-learning technologies and our ability to explain insights and learned topics.

Carla Gentry: My prediction -> 2020 is when some will finally realize, machine learning is a long and arduous journey that require LOTS of data and continuous tweaking (the end to “set it and forget it”?) – The most over hyped buzz of 2019 was AI, in order for it to be successful in 2020, the days of the “black box” has to be over! Explainable AI has got to be an essential component of a “Human AI”.  

Favio Vazquez: Data science is becoming a serious field. We will see an increase in important online and offline education about data science and its friends. Hopefully, we’ll become more confident in what we do and how we do it. Semantic technologies, decision intelligence, and knowledge data science will be our companions in the next years, so I recommend people to start exploring graph databases, ontologies, and knowledge representation systems.

Deborah Berebichez: The tedious and elementary steps of data science will be automated by various software; allowing data scientists to work on the more sophisticated aspects of data driven decision making. I think we are going through an uncertain time, economically and politically worldwide, and people are afraid because they are seeing a lot of change and the future seems uncertain. Most people, when they think of AI they are really thinking of what we call Artificial General Intelligence (AGI) which is the type of AI that can carry out any cognitive function in the way a human can. This technology is not here yet. Far from it. Today, what we have is called Narrow AI in applications such as IBM Watson, Siri, Alexa, and others. The main difference between AGI and Narrow AI is autonomy and objective-setting. That is, Siri only works if one asks it a question; it can’t automatically ask about the weather on its own. This doesn’t mean that AGI will not be here in the future. We have to accept that the world is changing without giving into fear. Of course, it’s natural to be afraid when we can no longer rely on the systems and the rules that we grew up with — especially the older generation. But don’t let fear lead you into the future, let curiosity lead you into it!

Lillian Pierson – As far as 2020 trends in AI, there are two areas that really catch my attention. Those are: (1) With improvements in facial recognition technologies, people and local governments will be pushing back even harder against the invasion of their privacy. (2) With 5G adoption steadily increasing, we’ll see more exciting initiatives arise at the area of convergence between 5G, IoT, and AI.

Rachel Stuve Anomaly detection will be a key trend in 2020. Detecting potential fraud, customer churn, medical claim denials, and purchasing overlaps (to name a few) is an area where AI will be leveraged to help organizations proactively address possible issues. Coupled with unique customer delivery (i.e. branded and differentiated user experience), this will be a key area that AI will make an impact in 2020.

Isaac Faber Ph.D. – Data Science is in the process of settling. Like the foundation of a house after it has been laid. There are two main camps; business-focused data scientist and developer-focused data scientist. The former has lost some of the spotlight to the later, mostly due to the deep-learning and machine learning hype. I think this will settle down a bit this year. Deep learning will be recognized as sometimes necessary but not sufficient for providing actual business value. The prominent players in the AI space will be people solving boring problems. The growing industry of robotic process automation (RPA) will likely, and quietly, capture most of the value. Probably the first few $B AI companies will be in the RPA, or similar, space.

Vin Vashishta – Deep Learning is going to run up against tough questions about bias. Corporate decision makers and government regulators are both going to be looking for details, not just promises and talk of debiasing.

Kevin Tran – I think the Forth Industrial Revolution is coming. Artificial intelligence is advancing and Google has claimed quantum supremacy. Quantum computing is an amazing breakthrough in computer science today. If this technology matures, it will make many infeasible complex computations feasible and that means our whole digital world will be transformed. An analogy to this would be upgrading from riding a horse to driving a car!

Kate Strachnyi – What we can expect to see in 2020 is the continued shift towards automating data analytics / data science tasks. Data scientists require tools that allow them to scale and solve more problems. This need will result in the development of automation tools across several stages of the data science process. For example, some data preparation and cleaning task are partially automated; however, they are difficult to fully automate due to the unique needs of companies. Additional candidates for automation include feature engineering, model selection, among others.

If you found value in this article, please share it with your network 🙂

 

Leave a Reply

Your email address will not be published. Required fields are marked *