CMS052 Abstract Data Types & Dynamic Data Structures Assessment 1
School of Computer Science
CMP3036M Data Science Page 1 of 2
CMP3036M Data Science Assessment 2 of 2 Criterion Grid 2016 – 2017
Learning Outcome Criterion Pass 2:2 2:1 1st
LO1 Critically
apply fundamental
concepts and
techniques in data
science.
LO2 Utilise state-
of-the-art tools to
design data science
applications for
various types of
data.
LO3 Analyse and
interpret large
datasets and
deliver appropriate
reports on them.
Section 1: Data
Summary, Pre-
processing and
Visualisation (5%)
This section is focused
on data summary, pre-
processing and
visualisation.
You have provided a
basic description of the
dataset, carried out some
data pre-processing steps,
and provided some plots.
Your discussion and
presentations are brief
with little critique.
You have described the
dataset to a good standard,
carried out some data pre-
processing steps, and
provided some plots. Your
discussion and
presentations are clear and
informative and provide
adequate detail on some
data pre-processing steps or
data visualisation or
interpretation.
You have described the dataset
to a very good standard,
carried out significant data
pre-processing steps, and
provided some plots to a good
standard. Your discussion and
presentations are clear and
detailed, with good focus on
the model development.
You have clearly described the
dataset to an excellent standard,
carried our substantial and key
data pre-processing steps, and
provided several interesting and
insightful plots. Your discussion
and presentations are detailed, in-
depth, and offer a critique of the
steps undertaken. A significant
amount of the discussion is
related to the model development.
Section 2: Comparison
of Algorithms (7.5%)
This section presents a
significant and detailed
discussion that covers
all main processes
carried out to compare
the required algorithms.
The following tasks have
been carried out to a basic
standard and may not be
fully complete: i) You
have split the data as
required, ii) You have
explained the MAE
metric, and iii) You have
reported (described) your
steps and your results
from at least one
algorithm as required.
The following tasks have
all been carried out to an
adequate, basic standard: i)
You have split the data as
required, ii) You have
explained the MAE metric,
and iii) You have reported
(described) your steps and
your results from at least
two algorithms as required.
Comparison of algorithms
is basic.
The following asks have all
been carried out to a good and
detailed standard: i) You have
split the data as required, ii)
You have explained the MAE
metric, and iii) You have
reported (described) your steps
and your results from at least
three algorithms as required.
Comparison of algorithms is
detailed with critique.
The following asks have all been
carried out to a very good and
detailed standard: i) You have
split the data as required, ii) You
have explained the MAE metric,
and iii) You have reported
(described) your steps and your
results from at least three
algorithms as required.
Comparison of algorithms is
significantly detailed with a
verbose critique.
Section 3: Evaluating
Model (15%)
This section should
discuss the process of
selecting and evaluating
the algorithms as
required.
You have set parameters
and tested the boosted
trees as required. You
have shown and
interpreted your results.
Your discussion and
presentation are brief and
basic. There are several
required steps missing.
You have set parameters
and tested the boosted trees
as required. You have
implemented the 10-fold
cross validation. You have
shown and interpreted your
results. Your discussion and
presentation are basic with
little depth.
You have set parameters and
tested the boosted trees as
required. You have
implemented the 10-fold cross
validation. You have shown
and interpreted your results
clearly. Your discussion and
presentation is significantly
detailed with good depth.
You have set parameters and
tested the boosted trees as
required. You have implemented
the 10-fold cross validation. You
have shown and interpreted your
results to a high standard that
demonstrates in-depth knowledge
and insight.
CMP3036M Data Science Page 2 of 2
Section 4: Time Series
Modelling (15%)
This section should
contain a detailed
discussion on your
solution of considering
time effect into your
predictive model
You have answered the
CEO’s questions and
concerns by repeating the
training of the linear
model and the regression
forests. Several other
required steps are missing
in your discussion.
You have answered the
CEO’s questions and
concerns by repeating the
training of the linear model
and the regression forests.
You have added 12 new
hour features to the data set
using Python or R code.
You have reported your
model performance.
Several other required steps
are missing in your
discussion.
You have answered the CEO’s
questions and concerns by
repeating the training of the
linear model and the
regression forests. You have
added the new hour and the
day features to the data set
using Python or R code. You
have reported your model
performance. Your discussion
on each step is clear but lacks
depth and critique.
You have answered the CEO’s
questions and concerns by
repeating the training of the linear
model and the regression forests.
You have added new hour and
day features to the data set using
Python or R code. You have
reported your model performance
to a high standard. Your
discussion on each step is clear,
detailed and demonstrates
excellent critique of the steps
carried out.
Section 5: Time Series
Prediction (7.5%)
This section should
contain a detailed
discussion on your
solution of predicting
for 2, 3, 4 and 5 hours
ahead.
You have shown your
method to predict for 2, 3,
4 and 5 hours ahead. Your
discussion and
presentation are brief,
incomplete or without
supporting evidence.
You have shown your
method to predict for 2, 3, 4
and 5 hours ahead. Your
discussion and presentation
are clear and informative.
You have shown your method
to predict for 2, 3, 4 and 5
hours ahead. Your discussion
and presentation are clear,
informative and in detail.
Discussion demonstrates a
good depth of understanding.
You have shown your method to
predict for 2, 3, 4 and 5 hours
ahead. Your discussion and
presentation are carried out to a
high standard, demonstrating
excellent critical understanding of
predicting.
Weighting Criteria in this assessment are weighted as indicated by the percentages presented above.