Skarp is an advanced predictive analytics engine that uses statistical algorithms and machine learning techniques to forecast the future.
It analyses historical internal data as well as external sources such as weather, holidays, economic indicators, footfall and more. This combination of advanced data science and multiple data sources is at the core of Skarp’s proposition, and why we can offer highly accurate forecasts at a fraction of the cost of doing it in-house.
Predictive analytics engine employs both Random Forest and Linear Regression algorithms
The best approach is selected from a single algorithm or a combination of multiple models
Underlying trends and seasonality within the source data are established at a granular level (e.g. daily/monthly/annual seasonality by location)
Machine learning algorithms run feature engineering on the raw data, including:
Non-linearity (e.g. the impact of temperature on sales)
Interaction (where two or more variables work in tandem)
Heuristics used to select significant features whilst avoiding ‘overfitting’
Model continually optimised as the predictive engine learns from new data
Multiple external data sources are integrated with our clients’ internal data to provide additional insight (at no extra cost)
This allows our analytics engine to identify the impact of everything from high street footfall to rainfall quantity or religious holidays.
There is no setup fee or minimum contract term with Skarp, and we offer all new clients a proof of concept free of charge. We believe the accuracy of our forecasts will speak for itself.