Reference no: EM132687457

BIG DATA AND DATA ANALYTICS

EXERCISE 1 [R-CODE]
Use R to create a variable called "posrating" in the dataframe. The variable takes on the value 1 if ratings >= 6.8. For ratings < 6.8 it takes on the value 0. Use R to perform a logistic regression that regresses the newly created variable "posrating" on aggregate_followers, dummy_sequel, netlikes, and sentiment in order to predict the probability that a movie has a positive rating. Interpret the coefficients and report the results of the logistic regression in APA style (including a logistic regression table and reporting of AIC-values).

EXERCISE 2 [R-CODE]
Use R to create a confusion matrix. Report the confusion matrix. Calculate and interpret sensitivity, specificity, and accuracy.

EXERCISE 3 [R-CODE]
Use R to create two logistic regression models. In the first model, regress posrating on sentiment. In the second model, regress posrating on budget, dummy_sequel, and sentiment. Plot an "Area under the ROC Curve" plot for each of the two models and explain the plots. Use the AUC values to compare the two models and interpret the results.

EXERCISE 4 [R-CODE]
Use the winsor function discussed in Week 3 to create a variable "sentiment_winsor" with a multiplier of 2.2. Use R to create a regression tree that uses budget and sentiment to predict ratings. Then, create a scatterplot of budget and sentiment, and add the partitions of the regression tree. Interpret at least 2 partions of the partitioned scatterplot.

EXERCISE 5 [R-CODE]
Based on the regression tree created in Exercise 4, use cross-validation to determine the optimal tree size and prune the tree. Plot and interpret the tree.

EXERCISE 6 [R-CODE]
Use R to create a classification tree to predict posrating. As predictors, take into account the variables aggregate_followers, comments, likes, dislikes, and sentiment. Use cross-validation to determine the optimal tree size and prune the tree. Plot and interpret the tree.

This lab project is based on a dataset about movie success in 2014 and 2015 by Ahmad et al. (2015) which is available on the online platform by Lichman et al (2013). Download the file movidata.csv from Blackboard and then practice the following topics in preparations for Lab Project 5.

PREPARATION
In preparation for Lab Project 5, load the moviedata.csv dataset. Use the dataset to practice the following topics:
- Logistic regressions: coefficient estimates, predicted probabilities, residuals, standard errors, confidence intervals, z-values, AIC, log likelihood
- Search for and watch online videos and blog entries about implementing a classifier (based on logistic regressions and classification trees; e.g., YouTube, R-Bloggers, Stackoverflow)
- Reporting of results in APA style
- Analysing and interpreting the results of logistic regressions
- Creating and interpreting a confusion matrix
- Sensitivity, specificity, and accuracy
- Area under the ROC
- Decision trees: regression trees and classification trees
- Pruning decision trees
- Analysing and interpreting the results of decision trees

Attachment:- Dataset_Description_and_Preparation.rar