An experienced data scientist can easily tackle a new data science problem and select the Machine Learning algorithms to solve the problem at hand. For new data scientists, however, selecting the best algorithm can be a problem by itself. Several companies and data scientists proposed guidelines for achieving the best results and they summarized them in cheat sheets built upon the rules they apply in their day-to-day work activities.
Scikit-Learn, the Pyhton library proposes its own cheat sheet with comprehensive rules for selecting the algorithms, even offering alternatives for the unfortunate situation when the first-choice algorithm is not working.
http://scikit-learn.org/stable/tutorial/machine_learning_map/index.html
SAS have their own take for a cheat sheet that depicts a diagram you should follow to select the needed algorithm making considerations upon the data dimensionality, data usage purpose, and urgency of the task.
https://blogs.sas.com/content/subconsciousmusings/2017/04/12/machine-learning-algorithm-use/
The Microsoft Azure cheat sheet comes in handy especially when designing your model in the platform they provide and is focused on predictive analytics tasks.
https://docs.microsoft.com/en-us/azure/machine-learning/studio/algorithm-cheat-sheet
The emoji Machine Learning cheat sheet
One of these cheat sheets caught my attention due to the use of emojis to help readers understand and remember the algorithm selection rules in a more appealing way. The data scientist who came up with this cheat sheet offers also details on how she chose the representation. Check it out here http://www.emilyinamillion.me/blog/2016/5/30/the-making-of-a-cheatsheet-emoji-edition
