Kotlin Coroutines – Flow parallel processing

Meant as an alternative to the Kotline Coroutine "Channel", a "Flow" is another way of enabling communication between two co-routines. The difference between the two is essentially that a channel is "hot" (i.e. starts as soon as it is declared), while a flow is "cold" (i.e. only starts when required (or "subscribed to" in reactive … Continue reading Kotlin Coroutines – Flow parallel processing

Machine Learning: Oversampling vs Sample Weighting

How do you "influence" a ML model? For example, imagine a scenario where you'd like to detect anomalies in a given data set. You reach for your favourite algorithm - in my case Isolation Forest: Our example output from Isolation Forest It does fine for most cases, except that one data point which invariably gets … Continue reading Machine Learning: Oversampling vs Sample Weighting