How AI Learns from Data

AI learns by finding patterns in data. Instead of being explicitly programmed with rules, an AI model is trained on examples. For instance, show an AI thousands of labeled pictures of cats and dogs, and it learns to distinguish them on its own. This process is called Machine Learning (ML).

• Training data: examples the model learns from (e.g., 10,000 labeled emails as "spam" or "not spam").
• Features: the pieces of information the model uses (e.g., words in an email, pixel values in an image).
• Labels: the correct answer the model is trying to predict (e.g., "spam" or "not spam").
• After training, the model can make predictions on new, unseen data — this is called inference.

💡 Practical Skill: Use Google Colab (free, browser-based) to run simple ML experiments. No installation needed — just a Google account. Start with pre-built datasets from scikit-learn.

# A simple ML classifier using scikit-learn# Run this in Google Colab or locally with: pip install scikit-learn from sklearn import tree # Features: [weight_grams, texture_smooth(0=rough,1=smooth)]# 0 = orange, 1 = applefeatures = [    [150, 0],  # orange    [170, 0],  # orange    [180, 1],  # apple    [200, 1],  # apple    [140, 0],  # orange    [190, 1],  # apple] # Labels: 0 = orange, 1 = applelabels = [0, 0, 1, 1, 0, 1] # Train the decision tree classifierclassifier = tree.DecisionTreeClassifier()classifier = classifier.fit(features, labels) # Predict a new fruit: 160g, smooth texture (should be apple)result = classifier.predict([[160, 1]])fruit = "Apple" if result[0] == 1 else "Orange"print(f"The AI predicts: {fruit}")# Output: The AI predicts: Apple