1. Briefly describe th emajor differences between data mining and statistics. a. Statistics is user driven, while data mining is data driven. b. In statistics, there exist underlying theory about certain relationships in data. While in data mining, there is often no pre-existing theory. c. In statistics, users use statistical methods to testify the hypothesis among data. While in data mining, users often use different techniques to examine data and uncover unknown relationships.

2. What can an organization do to deal with data problems such as missing data and outliers?
Missing data: a. Ignore the tuple b. Fill in the missing value manually c. Use a proxy variable with no missing values.
Outliers:
a. Delete rows. b. Recode c. Transform variables.

3. In a data mining exercise, a data set is usually partitioned into training, validation, and test data. Briefly describe the roles assumed by these partitions. a. Training data: used to build and fit models b. Validation data: used to monitor and fine-tune the model to improve its generalization. Tuning involves selecting competing models and optimize the selected model based on validation data. c. Test data: used to test the performance of model m unbiased assessment.

4. Which takes four possible values: freshman, sophomore, junior, and senior.
Recode the variable: replace freshman with 1, replace sophomore with 2, replace junior with 3, replace senior with 4. 5. Data cleansing a. I select the whole Comment table, and insert a pivot table in a new sheet. Then I summarize the Comment table b. I check each row of thread ID in pivot table to find out whether this specific thread ID is contained in Thread table by using Find c. I use red color to highlight the thread IDs which are not find in Thread table. d. I make a copy of Comment