Energy Appliances Dataset

The Dataset for this assignment is modified version of a subset of data used in Candanedo et al$,2017.$

The experimental data have been used to create models of energy use of appliances in a low$-$energy house.

The modified Dataset provides the energy use of Appliances $($denoted as Y$).$

The Dataset comprises $5$ features $($variables$),$ which are denoted as X$1,$ X$2,$ X$3,$ X$4$ and X$5.$

The details about these variables are given below:

X$1$: Temperature in living room area $($Celsius degrees$)$

X$2$: Humidity in living room area $($percentage$)$

X$3$: Temperature in office room $($Celsius degrees$)$

X$4$: Humidity in office room $($percentage$)$

X$5$: Pressure $($millimeter of mercury$)$

Y: Appliances energy consumption $($Wh$)$

For more information about the variables see Candanedo et al$,2017.$

Assignment tasks

T$1.$ Understand the data

$($i$)$ Download the txt file $($ENB$\_2023.$txt$)$ and save it to your R working directory.

$($ii$)$ Assign the data to a matrix, e$.$g$.$ using

the.data $<-$ as$.$matrix$($read$.$table$("$ENB$\_2023.$txt$"))$

$($iii$)$ The variable of interest is Y$.$ To investigate Y$,$ generate a subset of num$\_$row$=400($use the same setting

for the following tasks as well$)$ with numerical data e$.$g$.$ using:

my$.$data $<-$ the.data$[$sample$(1$:num$\_$samples,num$\_$row$)$ c$(1$:num$\_$col$)]$

This would give you a new dataset with num$\_$row rows and num$\_$col columns. Values of num$\_$sample and num$\_$col

have to be determined from the data provided.

$($iv$)$Use scatter plots and histograms to understand the relationship between each of the variables X$1$ X$2,$ X$3,$ X$4,$ X$5,$

and your variable of interest Y$,$ i$.$e$.,$ catter plots of $($X$1,$ Y$),($X$2,$ Y$),...,($X$5,$ Y$),$ and histograms of X$1$ X$2,$ X$3,$

X$4,$ X$5,$ Y$.$

T$2.$ Transform the data

Choose any FOUR variables from X$1,$ X$2,$ X$3,$ X$4,$ X$5.$

Make appropriate transformations so that the values can be aggregated in order to predict

the variable of interest Y$.$

Assign your transformed data along with your transformed variable of interest to an array

$($it should be $``$num$\_$row$$ rows and $5$ columns$).$ Save it to a txt file titled "name$-$transformed.txt$".$

write.table$($your$.$data,"name$-$transformed.txt$")$

The following tasks are based on the saved transformed data.

T$3.$ Build models and investigate the importance of each variable.

$($i$)$ Download the AggWaFit.R file to your working directory and load into the

R workspace using,

source$("$AggWaFit$718.$R$")$

$($ii$)$ Use the fitting functions to learn the parameters for

a$.$ A weighted arithmetic mean $($WAM$),$

b$.$ Weighted power means $($WPM$)$ with p $=0\mathrm{.}5,$

c$.$ Weighted power means $($WPM$)$ with p $=2,$

d$.$ An ordered weighted averaging function $($OWA$).$

T$4.$ Use your model for prediction.

Using your best fitting model from T$3,$ i$.$e$.,$ WAM, WPM$(0\mathrm{.}5),$ WPM$(2),$ or OWA, predict Y $($Appliances$)$

for the following inputs:

X$1=19\mathrm{.}1,$ X$2=43\mathrm{.}29,$ X$3=19\mathrm{.}7,$ X$4=43\mathrm{.}4,$ X$5=743\mathrm{.}6$

You should use the same pre$-$processing as in Task $2.$

Compare your prediction with the measured Y$=60.$

T$5.$ Summarise your data analysis in up to $20$ slides for a $5-$minute presentation

The slides should include the following content:

$-$ Correlations between the variables;

$-$ What kinds of data distributions you have identified in the raw data, use the histograms you have produced;

$-$ List and explain the transformations applied for the selected four variables and the variable of interest;

$-$ Explain the importance of the variables you have selected;

$-$ The best fitting model on your selected data; include two tables:

one with the error measures and correlation coefficients, and one summarizing the weights$/$parameters

and any other useful information learned for your data;

$-$ Your prediction result and comment on wheather you think it is reasonable;

$-$ Discuss the best conditions $($in terms of your chosen variables$)$ under which a low energy use of

appliances will occur.

$-$ Comment on the implications and limitations of the fitting model you used for prediction.