Why check first
Machine learning code is mostly ordinary data code with a small model step in the middle. If loading, cleaning and summarising data still feels hard, every lesson in this track will feel twice as hard, and you may decide wrongly that machine learning is not for you. Checking your readiness now saves you weeks.
Skill one: Python you can write without copying
You should be able to write a short Python script on your own: variables, lists, dictionaries, if statements, for loops and functions. You do not need to be an expert, but you should be able to read an error message and fix a simple bug.
Skill two: pandas and data cleaning
Almost every model starts as a pandas table. You should be able to load a CSV file, look at the first rows, select columns, filter rows, handle missing values and group data to get totals or averages. If df.groupby("region")["yield_kg"].mean() makes sense to you, you are in good shape.
Skill three: basic statistics
You should know what a mean, median and standard deviation tell you, what a sample is, and why a small or unfair sample can mislead you. Models are statistics at heart, and their results are always estimates with error.