In many of the AI projects Silo.AI has run, there’s one topic that creates some confusion: what is the difference between analytics – the stuff that’s been used for years in data science – and AI?
AI and machine learning have evolved from traditional analytics. The technologies often referred to as AI, such as machine learning, computer vision, and natural language processing, can complement existing analytical methods based on rules, optimisation, simulation, and statistical analysis.
In order for companies to leverage machine learning and AI, it is, however, crucial to understand where and when it makes sense to implement AI solutions, and when to leverage traditional analytics.
Relationship between analytics and AI
One way to think of AI is to see it as a set of four activities to achieve an outcome: sensing, reasoning, reacting and learning. When looking at AI from this perspective, the first activity, sensing, and the fourth activity, learning, are of particular interest as it is these two activities that differentiates AI from “traditional analytics”.
The two most prominent categories of “AI sensing” are computer vision (CV) and natural language processing (NLP). These technologies imply a new way of sensing data that previously could not be automatically processed, specifically relating to visual data in the form of images or video and textual data without any pre- defined logical structure, other than that of spoken language.
The learning aspect of AI implies the ability to adapt without specifically being reprogrammed. In contrast to, e.g., a rule-based system, a machine learning system is able to adapt independently when conditions change.
White Paper: AI & traditional analytics
To explore this topic in depth, we made a white paper that outlines a definition of artificial intelligence (AI) and machine learning and how these differ from traditional analytics.
After reading the white paper, you will understand:
1) What AI and machine learning means in practice
2) How using AI is different from rule-based systems, optimisation, simulation and statistical analysis
3) How to implement data driven decision making with machine learning, and
4) What type of use cases can be served with machine learning.
Get the White Paper on AI & traditional analytics.