Lift (data mining)

Key takeaways

Lift (data mining) belongs to science; place it in that map before memorising details.
Learn the definition first, then one example that makes the definition concrete.
Connect Lift (data mining) to a quantity you can measure, compute or draw — that is where exam questions come from.
Reproduce the core statement of Lift (data mining) from memory before moving on to harder problems.

Reference excerpt

In data mining and association rule learning, lift is a measure of the performance of a targeting model (association rule) at predicting or classifying cases as having an enhanced response (with respect to the population as a whole), measured against a random choice targeting model. A targeting model is doing a good job if the response P ( B ∣ T ) {\displaystyle P(B\mid T)} within the target ( T {\displaystyle T} ) is much better than the baseline ( P ( B ) {\displaystyle P(B)} ) average for the population as a whole. Lift is simply the ratio of these values: target response divided by average response. Mathematically,

lift = P ( B ∣ T ) P ( B ) = P ( T ∧ B ) P ( T ) P ( B ) {\displaystyle \operatorname {lift} ={\frac {P(B\mid T)}{P(B)}}={\frac {P(T\wedge B)}{P(T)P(B)}}}

For example, suppose a population has an average response rate of 5%, but a certain model (or rule) has identified a segment with a response rate of 20%. Then that segment would have a lift of 4.0 (20%/5%). Lift is also the ratio of precision and prevalence as shown in this chart:

Because lift is the ratio of precision and prevalence, and prevalence does not change with target set selection, lift is proportional to precision. Thus the lift curve, which plots lift vs recall (i.e., capture) is identical in shape to the precision-recall curve. The lift curve will always end at the point (1, 1) or (100%, 1) in the bottom-right corner, just as the precision-recall curve ends at the point (1/P(B), 1). This point corresponds to selecting the entire population as the target set. Like the precision-recall curve, the lift curve shows a similar tradeoff as the receiver operating characteristic (ROC) curve. It is also similar to the curve known in econometrics as the Lorenz or power curve.

Applications Typically, the modeller seeks to divide the population into quantiles, and rank the quantiles by lift. Organizations can then consider each quantile, and by weighing the predicted response rate (and associated financial benefit) against the cost, they can decide whether to market to that quantile or not.

Example Assume the data set being mined is:

where the antecedent is the input variable that we can control, and the consequent is the variable we are trying to predict. Real mining problems would typically have more complex antecedents, but usually focus on single-value consequents. Most mining algorithms would determine the following rules (targeting models):

Rule 1: A implies 0 Rule 2: B implies 1 because these are simply the most common patterns found in the data. A simple review of the above table should make these rules obvious. The support for Rule 1 is 3/7 because that is the number of items in the dataset in which the antecedent is A and the consequent 0. The support for Rule 2 is 2/7 because two of the seven records meet the antecedent of B and the consequent of 1. The supports can be written as:

supp ⁡ ( A ⇒ 0 ) = P ( A ∧ 0 ) = P ( A ) P ( 0 ∣ A ) = P ( 0 ) P ( A ∣ 0 ) {\displaystyle \operatorname {supp} (A\Rightarrow 0)=P(A\land 0)=P(A)P(0\mid A)=P(0)P(A\mid 0)}

supp ⁡ ( B ⇒ 1 ) = P ( B ∧ 1 ) = P ( B ) P ( 1 ∣ B ) = P ( 1 ) P ( B ∣ 1 ) {\displaystyle \operatorname {supp} (B\Rightarrow 1)=P(B\land 1)=P(B)P(1\mid B)=P(1)P(B\mid 1)}

The confidence for Rule 1 is 3/4 because three of the four records that meet the antecedent of A meet the consequent of 0. The confidence for Rule 2 is 2/3 because two of the three records that meet the antecedent of B meet the consequent of 1. The confidences can be written as:

conf ⁡ ( A ⇒ 0 ) = P ( 0 ∣ A ) {\displaystyle \operatorname {conf} (A\Rightarrow 0)=P(0\mid A)}

conf ⁡ ( B ⇒ 1 ) = P ( 1 ∣ B ) {\displaystyle \operatorname {conf} (B\Rightarrow 1)=P(1\mid B)}

Lift can be found by dividing the confidence by the unconditional probability of the consequent, or by dividing the support by the probability of the antecedent times the probability of the consequent, so:

The lift for Rule 1 is (3/4)/(4/7) = (3*7)/(4 * 4) = 21/16 ≈ 1.31 The lift for Rule 2 is (2/3)/(3/7) = (2*7)/(3 * 3) = 14/9 ≈ 1.56

… excerpt ends here. Continue reading the full article.

Worked examples

Example 1 — a first encounter with Lift (data mining)

Start with the simplest possible case. Write down what Lift (data mining) claims or describes in one sentence, then invent the smallest concrete situation in which that sentence is true. In science, the smallest case is usually a single object, a single equation or a single measurement. Check that every symbol or term in your sentence has a meaning in that case.

Example 2 — changing one variable

Take the situation from Example 1 and change exactly one quantity: double it, halve it, or set it to zero. Predict what should happen to Lift (data mining) before you calculate. Comparing your prediction with the result is the fastest way to find out whether you understand the idea or only the words.

Example 3 — an exam-style question

Typical questions about Lift (data mining) ask you to (a) state it precisely, (b) apply it to given data, and (c) explain a limitation. Practise writing all three answers in under five minutes; the third part is what separates a full-mark answer from an average one.

Applications of Lift (data mining)

In research: Lift (data mining) appears in science research whenever the underlying quantities have to be modelled precisely. Papers usually cite it as a starting assumption and then explore where it breaks down.
In technology and industry: Engineering practice reuses Lift (data mining) in design rules, simulations and safety margins. Knowing the idea lets you read a specification sheet and understand why the numbers look the way they do.
In the classroom: Lift (data mining) is common in secondary-school and first-year university syllabi. It links to neighbouring topics Data mining, so understanding it makes those chapters shorter.
In everyday life: Look for Lift (data mining) outside the textbook — in sport, cooking, traffic, electronics or the sky above you. An example you found yourself is remembered far longer than one you were given.

How to study Lift (data mining) in 20 minutes

Read the reference excerpt below once, without taking notes.
Close the page and write down what Lift (data mining) means in your own words.
Compare your version with the excerpt and mark what you missed.
Work through the three examples above with pen and paper.
Explain Lift (data mining) out loud to somebody else — or to Teacher Smith in the lgStudy chat.

Frequently asked questions

What is Lift (data mining) in simple terms?

In data mining and association rule learning, lift is a measure of the performance of a targeting model (association rule) at predicting or classifying cases as having an enhanced response (with respect to the population as a whole), measured against a random choice targeting model. A targeting mod…

Why does Lift (data mining) matter?

Because it connects several science ideas at once: it gives you a definition you can apply, a quantity you can calculate, and a way to check whether a result is plausible.

How should I study Lift (data mining)?

Read the excerpt, restate it from memory, then work through the examples and applications listed on this page. The five-step study plan above takes about twenty minutes.

What does this page cover?

It gives you a compact reference excerpt plus original lgStudy explanations, examples, applications and study material on Lift (data mining).