This dissertation was written as a part of the MSc in
Data Science
at the
International Hellenic University.
The accumulation and exploitation of data has brought about great variations in the
way people perceive their usefulness.
Actually
, there is an increasing interest in systematic data acquisition and the parallel extraction of useful information through them
in a
number of different sectors from medicine to sales
.
Data analysis
techniques can be applied to provide with predictions that could be utilised to the efficient
scheduling and
operation of
electricity generation.
The term smart city has been
around
for more than
two decades.
Nevertheless, it is increasingly enriched with new ideas and applications,
which mainly aim to provide citizens with a better standard
of living.
This dissertation focuses on the development and comparison of predictive algorithms under the smart city concept, utilising metered data on predefined time intervals.
More
specifically
, electricity consumption
as a total but also as main usage
s/spaces
breakdown
and weather data are
used to develop, train and test the models.
From a
technical point of view, a significant comparison
between different
machine learning
algorithms and methodologies is provided
. The outcomes
prove the
necessity of weather
data to predict residential
electrical consumption. Beside the fact that the available data
do not justify the term big data, the scalability of the model is examined in every step.
Collections
Show Collections
