Today I’m writing about relatively new Power BI visualization called Synoptic Panel. It allows to draw your own ”maps” from your own images and use those ”maps” to interact with your data. I’m using quotation marks around word ”maps”, because basically your image can describe whatever you want – a car, a warehouse, a human body or custom map. I think most typical use case would be using custom map, where reporting areas are not supported by Bing maps.

Jatka lukemista ”Custom maps/visuals in your Power BI report” →

This time I decided to write my blog in English, because this is a little bit deep s**t. I’m not describing how I created my machine learning project for predicting ice hockey results, because I have written about that in previous blogs, but I’m describing how I semi-automated the retraining process, which has to be done after every round.

First I needed those results and statistics from played games. I have described that process over here. In my training data I duplicate every game – one row for home team and one row for away team.

Jatka lukemista ”Azure Machine Learning retrain, running R scripts with Power BI and some DAX” →