AI and ML for Finance Practitioners

Topic 1: Introduction to Data Science & Big Data

Topic 2: Machine Learning: Introduction to Algorithms

Topic 3: Machine Learning: Regression,Support Vector Machine & Time Series Models

Topic 4: Machine Learning: Regularization, Regression Trees, Random Forest & Overfitting

Topic 5: Machine Learning: Classification & Clustering

Topic 6: Machine Learning: Performance Evaluation, Backtesting & False Discoveries

Topic 7: Data Mining & Machine Learning: Naïve Bayes & Text Mining

Topic 8: Big Data & Machine Learning: Ethical & Privacy Issues

Topic 9: Big Data & Machine Learning in the Financial Industry

Quiz LO 3.1.3

Test your knowledge of LO 3.1.3

17%

Question 1 of 6

1. Is the following definition of ‘decision surface of a classification tree’ generally correct?

Statement:
“The lines that separate the regions are known as decision lines (in two dimensions) or, in general, decision surfaces or decision boundaries (hyperplanes) in a multi-dimension space. Each node of a classification tree tests a single variable/feature against a fixed value so the decision boundary corresponding to it will always be perpendicular to the axis representing this variable/feature.”

Not correct

Correct

Question 1 of 6

Question 2 of 6

2. Which of the following statements describing the relationship between the decision surface and the number of variables is correct?

Statement I:
“Each node of a classification tree tests a single variable/feature against a fixed value, so the decision boundary corresponding to it will always be perpendicular to the axis representing this variable/feature. “
Statement II:
“In two dimensions, the line will be either horizontal or vertical. If the data had three variables, the instance space would be three-dimensional, and each boundary surface imposed by a classification tree would be a two-dimensional plane. “
Statement III:
“In higher dimensions, with n variables, each node of a classification tree imposes an (n-1)dimensional decision boundary (n-1 dimensional hyperplane) on the instance space.”

I and II

II

I, II, and III

None

Question 2 of 6

Question 3 of 6

3. Is the following statement about ‘frequency-based estimation of class membership probability’ generally correct?

Statement:
“The tree induction process divides the instance space into regions of class purity (low entropy). “

Not correct

Correct

Question 3 of 6

Question 4 of 6

4. Is the following statement generally correct?

Statement:
” If a leaf contains n positive instances and m negative instances, the probability of any new instance being positive may be estimated as $\boxed{P(\text { class positive })=\frac{n}{n+m}}$ “

Not correct

Correct

Question 4 of 6

Question 5 of 6

5. Is the following statement generally correct?

Statement:
” If a leaf contains few instances: n positive instances and m negative instances, the probability of any new instance being positive can be estimated with the Laplace correction defined by $\boxed{P_{\text {Laplace }}(\text { class positive })=\frac{n+1}{n+m+2}}$ “

Not correct

Correct

Question 5 of 6

Question 6 of 6

6. Assume that there are n=2 positive instances and m=0 negative instances, the probability of any new instance being positive is P(class positive) = 100% and negative P(class negative) =0. However, what would the probabilities of “class positive” and “class negative” be if you use the Laplace correction approach?

P(class positive) = 100% and P(class negative) = 0%

P(class positive) = 85% and P(class negative) = 15%

P(class positive) = 75% and P(class negative) = 25%

Question 6 of 6

Loading…

x

This Site Is Protected By
Shield Security →