We can all think of examples of successful forecasting, but what should we be aware of and what pointers can we adopt going forward?
Read moreDigitalization & IIoT
30 posts
Not all models need to be pre-trained. Sometimes it is more effective to apply algorithms inline to small batches of data.
Read more
The M-competitions compare and evaluate different approaches to, and implementations of, time-series forecasting. Here is a brief review of the latest one, M5
Read more
We must give consideration to the languages used in model training and model deployment – and we should do this before any model work begins. It is better to consider the two environments (and sometimes the two teams) as a whole and then work to a common interface.
Read more
Quality problems in the production of stamped and bent metal parts require data for analysis and sensors that provide it. Condition monitoring captures many data points that allow analysis and thus faster detection of the root causes of defects.
Read more
Manual, paper-based processes in intralogistics are inefficient, inaccurate and not scalable. RUCH NOVAPLAST, a manufacturer of product solutions made of particle foams, now uses RFID systems to track transport trolleys at every stage of production. This provides greater transparency in the manufacturing process.
Read more
Introduction Unsupervised anomaly detection with unlabeled data – is it possible to detect outliers when all we have is a set of uncommented, context-free signals? The short answer is, yes – this is the essence of how one deals with network intrusion, fraud, and other types of low-instance anomaly. In
Read more
Instead of randomly or exhaustively iterating through combinations of algorithms and parameters, we can use Bayesian Optimization libraries to build up an in-memory approximation to the process we want to fine-tune. We can then make a our selections on prior knowledge.
Read more
Time-to-event (TTE) use-cases crop up in many places across industries. Some examples would be: the prediction of customer churn (the sales domain), remaining-useful-life or time-to-failure TTF (predictive maintenance), or anomaly detection (machine monitoring). Some events are difficult to predict as they are hidden. We can instead try to look for interim events to improve prediction accuracy.
Read more
Time-to-event (TTE) use-cases crop up in many places across industries. Some examples would be: the prediction of customer churn (the sales domain), remaining-useful-life or time-to-failure TTF (predictive maintenance), or anomaly detection (machine monitoring).Some events are difficult to predict as they are hidden. We can instead try to look for interim events to improve prediction accuracy.
Read more
