Cognitive Insights Work Where Human Insight is Constrained
Cognitive insight applications are typically used where only a machine can really do the job, being exceedingly data intensive and involving high-speed number-crunching that would be beyond human capability. As such, they are unlikely to challenge human jobs, but provide a capability that was not possible previously.
These machine-learning applications are used to:
- predict behaviour – for example customer buying patterns
- identify fraud in real time
- analyse warranty claims for safety and quality issues in cars
- automate targeting of personalised advertising
- provide more accurate actuarial modelling
The difference from previous business analytics applications, which may either present historical data in user friendly form, lies in their ability to process massive volumes of data and derive patterns of data and insights that were not previously possible. Cognitive models:
- are extremely data intensive, often combined with ‘Big Data’ initiatives (what is the point of gathering large amounts of data otherwise?)
- are often trained on different aspects of the data set
- develop in their ability to learn over time from new data, observational feedback and external data feeds.
Is Machine Learning Becoming Artificial Intelligence?
Some versions of machine learning try to mimic humans in activities such as image or speech recognition, and whilst this is considered to lean towards true ‘artificial intelligence’, there is some way to go before such applications would be considered reliable enough to cater for the illogicalities as much as the logicalities of the human brain. There is much to do in the ‘real world’ before we get to the point where machines can be relied on to diagnose illnesses, for example, although they might well be able to prompt a medical professional to make diagnosis easier.
Cognitive Insight projects naturally involve circumstances where you don’t know what you don’t know and requirements can be very loosely defined, if at all. To get started with your project, or simply discuss your ideas, call us on +44 113 242 3795.