Questions $11-20$ are based on the files HW$7\_$training.csv and HW$7\_$scoring.csv Your goal is to predict the demand for product consumption. You will use linear regression to predict annual heating oil consumption and help a company to secure a certain amount of oil based on your predictions. You will build a model in the training dataset, and apply the results of your model in the scoring dataset.

$$

Question $11$

Upload the Training and Scoring datasets to RapidMiner. $($Note: Insulation is measured on a scale $1-10$ ten is the highest, outdoor temperature measured in F$,$ Num$\_$Occupants $$ number of total residents per home, Home$\_$Age $$ age in years, Home$\_$Size $$ size in sq ft$,$ Heating$\_$Oil$\_$Used $$ the number of units of oil purchased in a recent month$).$

$$

Run the process, and explore the results. In the Statistics view, check the ranges for each attribute in both training and scoring datasets. Note that the ranges for Home$\_$Size are different in the training and scoring datasets. Make the ranges identical by applying a filter in the Scoring dataset. Use Home$\_$Size $489$ and Home$\_$Size $7081.$

$$

Run the results. How many records are now in the Scoring dataset?

$7081$

$9436$

$9442$

$9988$

$5$ points $$

Question $12$

In the Training dataset, add operator Set Role, and assign the role of label to attribute Heating$\_$Oil$\_$Used. How many regular attributes are now in the dataset?

$3$

$5$

$7$

$9$

Question $13$

In the Training dataset, add operator Linear Regression, keep the defaults, run the process. Explore the Linear Regression Coefficients. Which attribute has the heaviest weight $($highest coefficient$)?$

Insulation$\_$Rating

Outdoor$\_$Temp

Home$\_$Age

Home$\_$Size

Num$\_$Occupants

$5$ points $$

Question $14$

In the Linear Regression Coefficients table, explore the significance of attributes. Which attribute is not significant and has been automatically removed from the model?

Insulation$\_$Rating

Outdoor$\_$Temp

Home$\_$Age

Home$\_$Size

Num$\_$Occupants

Question $15$

In the Linear Regression Coefficients table, explore the Intercept. What is the coefficient of the Intercept?

$210\mathrm{.}679$

$3\mathrm{.}329$

$2\mathrm{.}05$

$160\mathrm{.}634$

$5$ points $$

Question $16$

Based on the coefficients in the Linear Regression Coefficients table, create a Regression formula and calculate oil consumption for the house with the following attributes: Insulation$\_$Rating $=5,$ Outdoor$\_$Temp $=75,$ Home$\_$Age $=30,$ Home$\_$Size $=2200,$ Num$\_$Occupants $=5.$ What is the result of your calculation? Round up the number if necessary.

$180$

$189$

$201$

$220$

$5$ points $$

Question $17$

Based on a Regression formula created in Question $16,$ calculate oil consumption for the house with the following attributes: Insulation$\_$Rating $=9,$ Outdoor$\_$Temp $=45,$ Home$\_$Age $=15,$ Home$\_$Size $=1600,$ Num$\_$Occupants $=2.$ What is the result of your calculation? Round up the number if necessary.

$134$

$152$

$161$

$187$

Question $18$

Apply the Linear Regression model to the Scoring dataset. Run the process and explore the results. What is the average predicted oil consumption in the scoring dataset? Round up the answer if necessary.

$198\mathrm{.}7$

$218\mathrm{.}8$

$221\mathrm{.}5$

$304\mathrm{.}4$

$5$ points $$

Question $19$

Add operator Aggregate $($Note: connect the lab port of Apply Model and exa port of Aggregate$).$ In the Aggregate parameters, edit the list of aggregation attributes, and calculate the sum and median of the Prediction$($Heating$\_$Oil$\_$Used$).$ Based on your calculations, how many units of oil will be required to satisfy the demand? Round up the answer if necessary.

$1,587,999$

$1,865,345$

$1,989,868$

$2,064,756$

$5$ points $$

Question $20$

Based on the aggregate results, what is the median value for the Prediction$($Heating$\_$Oil$\_$Used$)?$ Round up the answer if necessary.

$198\mathrm{.}7$

$218\mathrm{.}8$

$221\mathrm{.}5$

$304\mathrm{.}4$