As an example of a new way to perform a root-cause analysis, here is an example in case of late invoice payments. Late invoice payments can worsen the relationship with your suppliers. That is why it is important to proactively manage problem areas where late payments often occur and to be able to take action.
For this analysis we make use of a rule-based Machine Learning algorithm that is able to predict a specific event (late payment) given a certain probability. Only features with high impact on the outcome are considered.
Ourlate payment algorithm provides the following benefits to your company:
For this analysis we make use of invoice data only. Data that our model requires to analyze late payments contain attributes like vendor name, invoice type, amount, payment date, employee, due date etc.
We train our model on your data and extract highly interpretable sentences pointing to certain problem areas. The output is provided through an interactive dashboard and static report containing all results.
Use our algorithm to learn highly interpretable patterns from your data that allows to keep focus on problem areas.
In the 'Late payment root cause analysis', the algorithm is able to detect for which cases and under which circumstances a late payment occurs.
In the first tab you will find the 9 most important insights with the most impact. These can be selected under 'Select a subset'. The insight determined by the algorithm is found as a text under 'insight'. The coherence and score per data element can be found under 'Data structure'.
In the second tab you will find the details of the various invoices that match the selected insight. You can visualize the individual processteps and there order by selecting a specific event within a case (within the table Events). Or you can visualize the whole sequence of all process steps graphically by selecting a specific case (within table Cases).
Note: all data-attributes and visuals in this webpage are based on fictitious data