All Matches

Solution Library

Expert Answer

Textbooks

Search Textbook questions, tutors and Books

Oops, something went wrong!

Change your search query and then try again

Result Not Found

Books FREE
Tutors
Study Help
Hire a Tutor
AI Study Help New

Search
Sign In
Register

study help
business
business analytics data

Questions and Answers of Business Analytics Data

Refer to the previous exercise. Apply the other two binning methods (equal width, and equal number of records) to this same variable. Compare the results and discuss the differences. Which method do
Report on whether anomalous fields exist in this data set, based on your EDA, which fields these are, and what we should do about it.
Why do we need to perform EDA? Why should not we simply proceed directly to the modeling phase and start applying our high-powered data mining software?
Apply the following inverse ln transformation to obtain the original value: eln value = exp(ln value).
Apply the following inverse Z transformation to obtain the original value: original value = (z value) ⋅ s + x. De-transforming a ln value:
Find the mean x and standard deviation s used to perform the standardization.
Report the standard errors (for continuous values) or confidence levels (for categorical values) for your imputations in Exercise 14
Impute all missing values in the data set. Explain the ordering that you are using.
Open the ClassifyRisk_Missing data set. Impute the missing value for marital status. Use the ClassifyRisk_Missing2 data set for Exercises 14–15.
Compare the standard errors for the imputations obtained in Exercises 9 and 11.Explain what you find.
Impute the sugars value of Quaker Oatmeal.
Impute the carbohydrates value of Quaker Oatmeal.
Impute the potassium content of Cream of Wheat.
Impute the potassium content of Almond Delight using multiple regression.
Repeat the procedure using the breast_cancer.arff data set with WEKA by selecting an attribute subset using Genetic Search. This time, however, specify naïve Bayes with use kernel estimator = true
(Extra credit). Write a computer program for a simple genetic algorithm. Implement the example discussed in the text, using the Normal (16, 4) fitness function. Let the crossover rate be 0.6 and the
Calculate the child vectors for the whole arithmetic crossover example in the text. Use the parents indicated in the section on simple arithmetic crossover, with ???? = 0.5. Comment on your results.
Continue the example in the text, where the fitness is determined by the Normal (16, 4) distribution. Proceed to the end of the third iteration. Suppress mutation, and perform crossover only once,
Compare the strengths and weaknesses of using backpropagation and genetic algorithms for optimization in neural networks.
Discuss why the selection operator should be careful to balance fitness with diversity. Describe the dangers of an overemphasis on each.
Match each of the following genetic algorithm terms with its definition or description. Term Definitiona. Selection One of the candidate solutions to the problem.b. Generation Scales the chromosome
Apply a misclassification cost of 5 (rather than the default of 1) for a false negative. Redo Exercises 23–29 using the new misclassification cost. Make sure to evaluate the models using the new
Evaluate all base classifiers, as well as the models defined by the candidate threshold values selected in the previous exercise, using overall error rate, sensitivity, specificity, proportion of
Scan the histogram from left to right, to identify candidate threshold values of the mean propensity for partitioning the test set into churners and non-churners. The goal is to select a set of
Construct a normalized histogram of mean propensity, with an overlay of Churn. (See Figure 26.3 for an illustration.)
For each record in the test data set, calculate the propensity of that record toward a positive response for Churn, for each of the base classifiers. Compute the mean propensity for each record
Use the training set to train a CART model, a logistic regression model, and a neural network model to be your set of base classifiers for predicting Churn.
Partition the data set into a training data set and a test data set.
Apply a misclassification cost of 2 (rather than the default of 1) for a false negative. Redo Exercises 23–29 using the new misclassification cost. Make sure to evaluate the models using the new
Evaluate all base classifier models and all voting ensemble models with respect to overall error rate, sensitivity, specificity, proportion of false positives, and proportion of false negatives.
Combine the classification results into voting ensemble models, using the following methods:a. Majority classificationb. Single sufficient classificationc. Twofold sufficient classificationd.
Apply the base classifier models to the test data set.
Use the training set to train a CART model, a logistic regression model, and a neural network model to be your set of base classifiers for predicting Income.
True or false: Ensemble models using voting or propensity averaging do not perform well with misclassification costs.
Describe how propensity averaging ensemble models would behave, for the following:a. Lower threshold valuesb. Higher threshold values.
How does a threshold value of t define positive and negative responses of the target variable?
When scanning the normalized histogram of mean propensity values, what should we look for in a candidate threshold value?
True or false: Propensity is a characteristic of a data set rather than a single record.
For an ensemble of m base classifiers, state in words the formula for mean propensity.
For a binary target, how is the propensity for a positive response calculated?
What is the rationale for using propensity averaging rather than a voting ensemble?
True or false: Voting ensemble models always perform better than any of their constituent classifiers.
Is a voting ensemble model constructed from the classification results of the training set or the test set?
What is a detriment of using voting ensemble models?
Describe the characteristics of the models associated with the following voting methods:a. Single sufficient classificationb. Positive unanimity classificationc. Majority classification.
Describe what negative unanimity would be.
Explain what single sufficient and twofold sufficient classification represent.
What is the difference between majority classification and plurality classification?
What is another term for simple model voting?
Calculate the pseudo-F statistic and p-value for each candidate, and select the candidate with the smallest p-value as the best clustering solution
Use a clustering algorithm to develop a clustering solution for a variety of values of k.
With the test data set, apply k-means with the value of k from the preferred model above. Perform validation of the clusters you uncovered with the training and test data sets of the preferred model.
Develop a good classification model for predicting loan approval, based solely on cluster membership. Apply data-driven misclassification costs as shown in Chapter 16. Compare your results for the k
Compare the pseudo-F statistics for the two cluster models. Which model is preferred?
Compare the mean silhouette values for the two cluster models. Which model is preferred?
Repeat Exercises 18–22 using k-means with k = 4.
Using the same variables as the previous exercise, provide a two-dimensional scatter plot, with an overlay of binned silhouette values, as shown in this chapter. Comment on the relationship between
Provide a two-dimensional scatter plot, using variables of your choice, with an overlay of cluster membership. Choose variables that result in an interesting plot. Note where the cluster boundaries
Calculate the mean silhouette values for each cluster, as well as the overall mean silhouette for the cluster model.
Generate a silhouette plot of your cluster model.
Use k-means with k = 3 to generate a cluster model with the training data set.
What are the criteria for determining whether there is a match between the clusters uncovered in the training and test data sets?
Why might statistical hypothesis tests not be very helpful for big data applications?
What is our cluster validation methodology?
True or false: The best clustering model is the one with the largest value of pseudo-F. Explain.
Explain how we can use the pseudo-F statistic to select the optimal number of clusters.
Why does the pseudo-F statistic have the word pseudo in its name?
Explain how the pseudo-F statistic accounts for both separation and cohesion.
Should the analyst always choose the cluster solution with the better mean silhouette value? Explain.
Describe what a silhouette plot is.
When will a data value have a perfect silhouette value? What is this value?
How is average silhouette interpreted?
Explain how silhouette accounts for both separation and cohesion.
How do we interpret a silhouette value?
What is a silhouette? What is its range? Is it a characteristic of a cluster, a variable, or a data value?
Why is SSE not necessarily a good measure of cluster quality?
What is cluster separation and cluster cohesion?
Why do we need evaluation measures for cluster algorithms?
Briefly profile the clusters for the winning model from the previous exercise.
Calculate model cost for each of the five different sortings. Which model has the highest profitability or the lowest cost?
Run BIRCH on each of the five different sortings. Report the value of k and the MS for each.
Generate four different sortings of the Loans_training data set. Together with the original order from the No Interest model you generated earlier, this makes five different sortings.
Based on your work in the previous exercises, what is the lesson we should learn? For Exercises 13–16, using the Loans data set, demonstrate that different sortings may lead to different numbers of
Evaluate each CART model using the Loans_test data set. Provide contingency tables. Compare the model costs, as in Table 21.4.
Using the Loans_training data set, develop CART models for predicting loan approval, based on cluster membership only, for the two cluster models.
Follow the methodology in Case Study Lesson One to develop cluster models with and without interest.
Is the MS value always indicative of the best cluster solution?
Why is it bad practice to include two highly correlated inputs to a clustering algorithm?
Why is Phase 2 of the BIRCH algorithm efficient?
Describe the parameters of the CF tree.
How are the CFs for two clusters merged?
What is a CF?
Describe the two phases of the BIRCH clustering algorithm.
Why is BIRCH appropriate for streaming data?
Use cluster membership as a further input to a C4.5 decision tree model for classifying income. How important is clustering membership in classifying income? Compare to the CART model.
Use cluster membership as a further input to a CART decision tree model for classifying income. How important is clustering membership in classifying income?
Using the information above and any other information you can bring to bear, construct detailed and informative cluster profiles, complete with titles.
Generate numerical summaries for the clusters. For example, generate a cluster mean summary.
If your software supports this, construct a web graph of income, marital status, and the other categorical variables. Fine-tune the web graph so that it conveys good information.
Construct a bar chart of the cluster membership, with an overlay of marital status. Discuss your findings.

Showing 900 - 1000 of 2834

First
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
Last

Services

Sitemap
Fun
Definitions
Become Tutor
Study Help Categories
Recent Questions
Expert Questions

Company Info

Security
Copyrights
Privacy Policy
Terms & Conditions
SolutionInn Fee
Scholarship

Get In Touch

About Us
Contact Us
Career
Jobs
FAQ
Campus Ambassador

Secure Payment

Download Our App

© 2024 SolutionInn. All Rights Reserved