When used from a research and strategic perspective, machine learning takes on a new guize. Whereas behavioural or transactional data alone produces significant gains in prediction, applying machine learning to survey data that captures the psychology of consumers can take machine learning to the next level!
We have been applying machine learning methods to survey data for over a decade and have produced significant gains in prediction. We have also used machine learning exclusively to tag research data to customer data.
Feature engineering is the process of creating new data (variables) out of existing factors that might not be immediately obvious, due to domain knowledge. A very simple example is when salutation may denote marital status and age or even income potential, or members of a household denoting household expenditure or the likelihood of insurance needs.
The ability to extract new information from existing information is both difficult and complex, but it does provide machine learning models with the potential to outperform traditional modelling. We are able to bring to life dormant or otherwise non-predictive factors by fusing attitudes, interests, opinions and needs to these factors to increase the predictive capability of your data!
Psychological epistemology is a discipline in psychology especially in decision sciences that attempts to account for ideas, actions, feelings, knowledge social interactions and underlies nudge marketing.
As we are a full market research agency, via the use of latent (hidden) structure and probability models associated with the likelihood of particular and specific human judgement occurring in real markets, we are able to create information and data that represents how consumers view the world. This information is then utilised by the machine learning models that we build in order to optimize performance and accuracy so your marketing communications hit targets more often!