03 Classification

Question 1

What does random_state?

Answer 1

it produces reproducible result

Question 2

Does shuffling training data improve model performance

Answer 2

Shuffle training set as some models perform poorly when instances are ordered.

Question 3

What is k-fold cross validation

Answer 3

It divides the dataset into k parts and model is trained on k-1 parts and validated on the kth part.
These parts get shuffled on each iteration.

Question 4

What is true positive and true negative

Answer 4

True positive- their actual value is 1 and model also predicted as 1.
True negative- their actual value is 0 and model also predicted as 0.

Question 5

What is false positive and false negative

Answer 5

False positive- their actual value is 0 but model predicted them as 1.
False negative- their actual value is 1 but model predicted them as 0.

Question 6

Confusion matrix format

Answer 6

T.P F.N
F.P. T.N

Question 7

Define accuracy

Answer 7

% of correct prediction made by our model
Formula- (T.N+T.P)/(TP+TN+FP+FN)

Question 8

When to use accuracy

Answer 8

It is best to use when there is class balance and worst to use when there is class imbalance.

Question 9

Define precision

Answer 9

Among all the positive PREDICTIONS how many are actually positive.
Formula- TP/(TP+FP)

Question 10

Define recall

Answer 10

AKA Sensitivity
Among all the ACTUAL positive’s how many are correct
Formula- TP/(TP+FN)

Question 11

When to use Precision

Answer 11

When our objective is to minimize false positives
e.g. - if we want our model to catch criminals in this case we let go of some criminal but can not catch an innocent person hence we need to reduce false positives.

Question 12

When to use recall

Answer 12

when our objective is to reduce false negatives
e.g. - suppose we want our model for intense checking in airport check-in. in this case it is ok to take an innocent person aside as we can check and let him go but we can not let go of a criminal hence we need to reduce false negative

Question 13

what relation between Recall & Precision

Answer 13

Inversely Proportional.

Question 14

When to use F1-Score

Answer 14

When we cannot trade of between false positives and false negatives
e.g.- we want a model to predict promotion of an employee in this case we dont want to stop promotion of a deserving employee we also dont want to promote someone not good hence we need both false positive and false negative

Question 15

Define F1-Score

Answer 15

It is harmonic mean of precision and recall (we choose HM cause in HM even if either of precision or recall goes low the value reduces drastically)

Question 16

Function used to get any score.

Answer 16

cross_val_score(sgdclassifier, x_train, y_train, cv = 3, scoring = “accuracy”)

Question 17

Impact on threshold on Precision and recall

Answer 17

When we increase the threshold the precision increases and the recall decreases.
As the threshold is decreased the recall increases and the precision decreases.

Question 18

How can we view decision score?

Answer 18

using decision_function()
we can not view the threshold but we can see the scores.

Question 19

ROC - AUC Curve

Answer 19

Plot between True Positive Rate (TPR) & False Positive Rate (FPR)

Question 20

Define True Positive Rate (TPR)

Answer 20

Recall AKA Sensitivity

Question 21

Define False Positive Rate (FPR)

Answer 21

is the number of negative instances which were wrongly identified as positive.
formula: 1-TNR(True Negative Rate i.e. negative instances which were correctly identified as negative AKA Specificity)

Question 22

which function to use to get probabilities of each instance

Answer 22

predict_proba()

Question 23

What is multiclass classification

Answer 23

it distinguishes between multiple classes

Question 24

which model are capable of classifying multilabel classification

Answer 24

Random Forest & Bayes Classifier

Question 25

Name few strictly binary models

Answer 25

SVM & Linear Classifiers

Question 26

How to use a binary classification model for multiclassification model

Answer 26

1. One versus all
2. Ove versus one

Question 27

Define One versus all strategy

Answer 27

in this if we want to classify 0-9 digits then we will build 10 classification models and we will consider score which is highest in each model.

Question 28

which among one versus all and one versus one strategy is preferred

Answer 28

One versus all is preferred.
Scikit-learn also uses this model by default for all binary classification

Question 29

Explain One versus One strategy

Answer 29

in this we will build one classification model for each pair.

Question 30

Best model to start with

Answer 30

Stochastic Gradient Descent
Especially with large datasets

Question 31

What are Multi label classification

Answer 31

e.g. - suppose i need a model to recognise me, jb and pj.
now in one instance i have a picture of me and jb then i need the model to give a output as 1,1,0 that is a multi label classification model