ASSESSMENT TASK 2 (PROBLEM SOLVING) in 2023T3
Using aggregation functions for data analysis
The provided zip file contains the data file [ENB_2023.txt] and the R code [AggWaFit718.R]
to use with the following tasks, include these in your R working directory.
Total Marks 100, Weighting 20%
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 X1, X2, X3, X4 and X5.
The details about these variables are given below:
X1: Temperature in living room area (Celsius degrees)
X2: Humidity in living room area (percentage)
X3: Temperature in office room (Celsius degrees)
X4: Humidity in office room (percentage)
X5: Pressure (millimeter of mercury)
Y: Appliances energy consumption (Wh)
For more information about the variables see Candanedo et al, 2017.
Assignment tasks
T1. Understand the data
(i) Download the txt file (ENB_2023.txt) from CloudDeakin 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 X1 X2, X3, X4, X5,
and your variable of interest Y, i.e., catter plots of (X1, Y), (X2, Y), …, (X5, Y), and histograms of X1 X2, X3,
X4, X5, Y.
T2. Transform the data
Choose any FOUR variables from X1, X2, X3, X4, X5.
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.
T3. 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,
(ii)
a.
b.
C.
d.
source("AggWaFit718.R")
Use the fitting functions to learn the parameters for
A weighted arithmetic mean (WAM),
Weighted power means (WPM) with p = 0.5,
Weighted power means (WPM) with p = 2,
An ordered weighted averaging function (OWA).
T4. Use your model for prediction.
Using your best fitting model from T3, i.e., WAM, WPM(0.5), WPM(2), or OWA, predict Y (Appliances)
for the following inputs:
X1= 19.1, X2=43.29, X3=19.7, X4=43.4, X5=743.6
You should use the same pre-processing as in Task 2.
Compare your prediction with the measured Y=60.
T5. 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. The slides should contain all necessary information to prove your findings. All the bold terms above must
appear in slide titles. For the 5-minute presentation, you may provide a link to YouTube or upload a mp4 video.
Any content beyond 5 minutes will not be graded.
SUBMISSION:
Submit to the SIT718 CloudDeakin Dropbox.
Your final submission must include the following TWO files:
1. The presentation slides with video, "name-slides" (pdf), covering all of the items in above
(where "name" is replaced with your name -you can use your surname or first name)
(a link to YouTube or uploading a mp4 file).
2. The R code file (that you have written to produce your results) named "name-code.R" (where "name" is
replaced with your surname or first name; .RMD file is not allowed).
Your assignment will not be assessed if the code is missing, or the outputs of the code are inconsistent
with the content of the slides.
For referencing, follow the Harvard style:
https://www.deakin.edu.au/students/studying/study-support/referencing/harvard
You must cite all the datasets, packages and literature you used for this assessment.
You will loose some marks for lack of or inappropriate citations/references.
References
Luis M. Candanedo, Veronique Feldheim, Dominique Deramaix. Data driven prediction models of
energy use of appliances in a low-energy house, Energy and Buildings, Volume 140, 1 April 2017,
pages 81-97, ISSN 0378-7788.
The original data are available in:
http://archive.ics.uci.edu/ml/datasets/Appliances+energy+prediction
