How to choose the right algorithm?

CarlosCV report abuse

I have finished building a pipeline for my DS project (binary classification). Results after running the pipeline show that the best accuracy have both XGBoost and LightGBM algorithms. Which one should I choose? Or which additional metrics can help me to decide which is the best?


CarlosCV report abuse
Any advice? Or perhaps do you know a good article :)?


communitymanager report abuse

Hi! Probably you won't get an answer because you will find many suggestions and articles if you enter your question directly into Google. Try it out and you will quickly find what you are looking for. Good luck and if you have any more detailed questions, please feel free to ask again!

Marjo3d report abuse

There are several different metrics for classification. Generally, accuracy is considered as a not very reliable metric. You can choose F1 score, precision, recall, ROC AUC metrics for example. It is often enough to look at accuracy, F1 score, and ROC AUC. Also, building a confusion matrix could help.

Add Answer

Need support?

Just drop us an email to ... Show more