It may be complicated to get access to huge amount of data. Yet, is this the only challenge? Surely not. As it is well-known in the machine learning community, having data is not tantamount to having knowledge. The Analytics Toolbox simplifies the extraction of the latter, by providing a set of libraries and modules designed to satisfy most data-related needs, and based on the most recent concepts and algorithms developed by the scientific community. It is buttressed by a modular structure, in which new analytics services can be added to fulfil new requirements; and in which multiple algorithms can be chained together, to give answer to even more complex questions.

But how will this Analytics Toolbox help you?

​

• By enabling fast prototyping. No data to download, no library to develop and deploy in-house. The Analytics Toolbox enables performing a first feasibility evaluation of a new business idea at essentially no cost.

​​

• By unleashing the power of advanced algorithms. The Toolbox includes modules not easily available in other all-purpose analytics solutions, and specifically designed with CPP data in mind. These span from the analysis of thousands of trajectories, to the representation of network relationships. Again, no in-house development is required: the Toolbox includes everything is needed for a first evaluation.

​​

• By minimising overheads. Filter your data prior to download, for instance through averaging, clustering, or through event-driven triggers. Only download what you need, and when you need it.

​

Which analyses can be executed?

• Basic statistics and data aggregation: from statistical metrics of location and dispersion, to analysis of distributions and entropies.

• Time series analyses, including the detection of drifts, or sudden changes, through the application of statistical and data mining models.

• Trajectory analyses. From the processing of individual trajectories, including interpolations and error detections; to multivariate scenarios, as in the detection of clusters of similar trajectories - see the image for an example.

• Network analyses, for understanding the structures created by interacting elements.

Read more: