Prediction

Question 1

Classification

Answer 1

Predicting when y=1 vs 0 (like churn). Tells you probability of y=1

Question 2

Unconditional Probability

Answer 2

Unaffected by previous or future events

Question 3

Law of Large Numbers

Answer 3

Increased observations means increased precision of prediction

Question 5

Conditional Probability

Answer 5

Stronger prediction, rather than just 25% of customers churn you divide churn rate by senior and not senior

Question 6

Bayes Rule

Answer 6

P(A|B) = P(A&B)/P(B)

P(A|B) = P(A)P(B)/P(B)

Question 7

Threshold

Answer 7

Value between 0 and 1. The value for probability that will mean it is considered a 1

Question 8

Sigmoid Function

Answer 8

H= @ + @ * tenure and then L(H) = 1/(1+e^-H)

gets you those predictions on multiple variables

Question 9

Certainty through |H|

Answer 9

|H|> 2 okay
|H|> 5 quite sure
|H| > 10 super sure

Question 10

Scatter Plotting Continuous Features

Answer 10

A method to analyze how your y value is affected by both. have feature 1 on the x, feature 2 on the y, and then two colours for whether each data point is y=1

Question 11

Precision

Answer 11

How many false positives, if higher than less Fp

Tp/(Tp + Fp)

Question 12

Accuracy

Answer 12

How many correct predictions (so true positives and true negatives)

(Tp + Tn)/(all obs)

Question 13

Recall

Answer 13

How many false negatives, if higher than less Fn

Tp/(Tp+Fn)

Question 14

F1

Answer 14

Balance between precision and recall

2((PR)/(P + R))

Question 15

Underfitting

Answer 15

Model bad in testing, not flexible enough

High bias

Add observations or add features

Question 16

Overfitting

Answer 16

Poor predictions in test data, too flexible on training data, too custom

High variance

less features or regularization (make simpler) or increased observations

Make sure 70/30 data split

Question 17

What to do when y= 0,1,or 2

Answer 17

Make z0=1 only if y= 0 then do Pr(z0|X)

Do the same with a z1 and z2

Question 18

Supervised Learning

Answer 18

Predict a labelled attribute y with features

classification, log regression

ex. home sales price (cond mean at each value to get line)

Question 19

Unsupervised Learning

Answer 19

no labelled target attribute to predict

come up with cluster observation on similar features

ex. market segments, similar images, news article types

Question 20

Anomaly Detection

Answer 20

Evaluate whether certain data anomalous

fraud detect, defect detect

Question 21

Reinforcement Learning

Answer 21

Implements actions, generates data, updates algorithm

Trade off between exploit and explore

Recco systems

Question 22

Machine Learning Benefits and Types

Answer 22

More precise because it can use more obs, more flexible forms, more variables

predicts then decides off that

Supervised Learning
Unsupervised Learning
Anomaly Detection
Reinforcement Learning

Question 23

External Validity

Answer 23

Model trained in context A can be used in context B

Likely higher when data context is similar