Data analytics is not only being revolutionised by artificial intelligence (AI) but cannot reach its full potential without the emerging technology, according to a new report.
GlobalData’s Data Analytics report notes that the data analytics market is relatively mature but that capabilities for prescriptive analytics – the most advanced of the four primary types of data analytics – are still maturing.
In increasing levels of sophistication, descriptive analytics simply summarises what things have happened based on analysing datasets, diagnostic analytics answers questions about why things have happened, predictive analytics suggests what might happen in the future and prescriptive analytics provides recommendations to companies of what to do next.
The report explains: “Prescriptive analytics, often described as the last mile of business analytics, uses AI to process historical data, uncover unseen patterns and relationships, create models to show the likelihood of scenarios or outcomes and then advise a course of action based on the findings. Complex decision-making is optimised using millions of decision variables, constraints and trade-offs.
“Using rule-based approaches and optimisation techniques helps decision-makers by providing recommendations and optimal actions to reach business goals and stops people from relying on their intuition. Applications include demand forecasting, supply chain logistics, power generation and employee scheduling.”
By way of example, the report outlines how food suppliers can combine historical data with forecasts (about upcoming events or the weather, for example) to proactively respond to changing demand across supply chains and ensure that the right products are in stores when needed.
The type and volume of calculations required to make such predictions necessitate sophisticated AI algorithms and significant computing power to facilitate them.
Of the importance of AI for data analytics, the report explains: “Without applying AI, particularly machine learning and generative AI technologies, to data analysis, realising its full potential is simply not feasible.
“One of the five V’s of big data is velocity. For data insights to be actionable and valuable, they must come quickly. Analytics processes must be self-optimised and able to learn from experience regularly, an outcome that is only possible with AI functionality and modern database technologies.”
GlobalData estimates the total data analytics market to have been worth $112.3bn in 2023 and expects it to hit $190bn by 2028, with a compound annual rate (CAGR) across that period of 11.1%.
link