Machine Learning (ML) · Beginner level

Beginner level · Module 1: Machine Learning Foundations · Lesson 3 of 3

Notebooks, pandas and scikit-learn

Set up the tools, load a data set with pandas and train a first model with scikit-learn's fit and predict.

In this lesson you will

  • Use a Jupyter or Colab notebook
  • Load data with pandas and train a model with fit and predict

Key concepts

notebookDataFrameestimatorfit and predict

The lesson

Notebooks mix code, results and notes, which suits exploring data. Google Colab runs in a browser with nothing to install; Jupyter runs locally through Anaconda or pip and works offline.

pandas loads tables into DataFrames: df = pd.read_csv("plots.csv"); df.head() shows the first rows and df.describe() summarises the numbers.

scikit-learn gives every model the same interface: create an estimator, call model.fit(X, y) with features X and labels y, then model.predict(X_new). Learning this pattern once lets you try many algorithms with almost no code changes.

Step by step

  1. Open a notebook in Colab or Jupyter.
  2. Load a small CSV with pandas and inspect it.
  3. Split it into features X and label y.
  4. Fit a simple model and make a prediction.

Practice activity

Load the provided crop-plot data set, train a demonstration model and predict the best crop for two new plots.

Expected result: The notebook runs top to bottom and prints predictions for the two plots.

Check your understanding

  1. Which two methods does every scikit-learn model provide?

    Show answer

    fit and predict.

  2. What does df.describe() show?

    Show answer

    Summary statistics for the numeric columns.

Taught live by an instructor — online or in person. Free course lessons are for self-study and do not lead to a certificate — see registration options.