Sasin MiM · Supervised learning
Run a machine learning project step by step, on our datasets or your own file. The computer does the maths; you make the decisions.
A dataset is a table: each row is one observation, each column is one piece of information. The target is the column you want to predict.
Before the model sees anything, lock some rows away. The model learns from the training set. The test set is used only at the end, to check the model on rows it has never seen.
Features are the columns the model may use as input to predict the target. Every feature must be known at the moment you make the prediction. Nothing is ticked for you: decide which columns make sense, and why.
The algorithm is the method that learns from the training set. Training produces a model: a rule the computer can apply to new rows.
How well does the model do on rows it has never seen? Compare it with a baseline: what you would get with no model at all.
Every model you train is logged here so you can compare experiments. The log disappears when you close or reload the page.