NORMA eResearch @NCI Library

OPTUNA Optimization Based CNN-LSTM Model for Predicting Electric Power Consumption

Ekundayo, Ibidokun (2020) OPTUNA Optimization Based CNN-LSTM Model for Predicting Electric Power Consumption. Masters thesis, Dublin, National College of Ireland.

[thumbnail of Master of Science]
PDF (Master of Science)
Download (2MB) | Preview


Forecasting residential energy consumption using deep neural networks has been attempted in past researches. Typically, optimizing these networks relies on the operator’s prior knowledge. They are also affected by the size of the search space and the tuning parameters for the model. In this research, we integrate OPTUNA, an optimization algorithm that can automatically determine different frames of hyperparameters to tune a CNN-LSTM neural network for predicting electric power energy consumption. Our research findings indicate that the proposed model can be applied as a potential alternative forecasting framework for better forecast accuracy and a broader generalization potential. The OPTUNA hyperparameter can remove mutation operations and crossover in comparison to the traditional models. To validate the potential of our proposed algorithm, we have selected the household electric power consumption dataset from the UCI machine learning public repository. Our proposed OPTUNA CNN-LSTM model explores varying degrees of optimum forecasting frameworks automatically and presented a lower MSE compared to a result of past literature that implemented the particle swarm algorithm.

Item Type: Thesis (Masters)
Subjects: Q Science > QA Mathematics > Electronic computers. Computer science
T Technology > T Technology (General) > Information Technology > Electronic computers. Computer science
Q Science > QA Mathematics > Computer software
T Technology > T Technology (General) > Information Technology > Computer software
Divisions: School of Computing > Master of Science in Data Analytics
Depositing User: Dan English
Date Deposited: 22 Jan 2021 11:43
Last Modified: 20 May 2021 10:59

Actions (login required)

View Item View Item