Q.Some flowers selected for artificial hybridisation do not require emasculation but bagging is essential for them. Give reason.

Let’s begin with something you already know from everyday life.

Imagine you have to make a very important decision — say, which college course to pick. You could ask one friend for advice. But that one friend might be biased (they love their own subject) or misinformed (they heard something wrong). So instead, you ask ten different people — teachers, seniors, parents, counsellors — and then take a majority vote or an average of their opinions. That collective decision is almost always more reliable than any single person’s advice.

That is the core intuition behind Bagging.

What Bagging Does

In data analysis and machine learning, Bagging stands for Bootstrap Aggregating. It is a technique used to make a prediction model more stable and accurate.

Here is the precise meaning: instead of building one single model (like one decision tree or one rule-based classifier) on the entire dataset, you create many slightly different versions of the same model, each trained on a different random sample of the original data. Then you combine all their outputs — usually by taking a majority vote (for classification) or an average (for prediction).

The key word is Bootstrap. A bootstrap sample is created by randomly picking data points from the original dataset with replacement — meaning the same data point can be picked more than once in a single sample, and some points may be left out. Each sample is different, so each model sees a slightly different view of the world.

Why It Matters

The main utility of Bagging is reducing variance — that is, reducing the model’s tendency to change wildly if you give it a slightly different set of data.

Think of it this way: a single model can be “nervous” — it might overreact to a few unusual data points (outliers) or to random noise in the data. But when you average the opinions of many such models, their individual nervousness cancels out. The combined model becomes calm, stable, and more accurate.

Key Points to Remember (for exams)

• Purpose: To improve accuracy and stability by combining multiple models.
• How it works: Create many random samples (with replacement) from the original data. Train a separate model on each sample. Combine their predictions.
• Why it helps: Reduces overfitting (the model memorising noise) and lowers variance.
• Trade-off: It does not reduce bias much — if the individual model is fundamentally weak, bagging won’t fix that. It only makes a good model more reliable. …