Predicting electricity energy consumption: A comparison of regression analysis, decision tree and neural networks

Tso, Kwok Fai; Yau, Kelvin K. W.

doi:10.1016/j.energy.2006.11.010

Cited by 814 publications

(350 citation statements)

References 11 publications

Supporting

Mentioning

301

Contrasting

Unclassified

Order By: Relevance

“…In addition to this, it is much faster in speed and is more effi cient than C4.5 in terms of memory usage. According to (Tso and Yau 2007), C5 has the following advantages over C4.5 algorithm: "(1) the branch-merging option for nominal splits is the default; (2) misclassifi cation costs can be specifi ed; (3) boosting and cross-validation are available; and (4) the algorithm for creating rule sets from trees is much improved".…”

Section: Artifi Cial Intelligence Techniquesmentioning

confidence: 99%

A Comparison of Data Mining Techniques for Credit Scoring in Banking: A Managerial Perspective

İnce

Aktan

2009

Journal of Business Economics and Management

View full text Add to dashboard Cite

Abstract. Credit scoring is a very important task for lenders to evaluate the loan applications they receive from consumers as well as for insurance companies, which use scoring systems today to evaluate new policyholders and the risks these prospective customers might present to the insurer. Credit scoring systems are used to model the potential risk of loan applications, which have the advantage of being able to handle a large volume of credit applications quickly with minimal labour, thus reducing operating costs, and they may be an effective substitute for the use of judgment among inexperienced loan offi cers, thus helping to control bad debt losses. This study explores the performance of credit scoring models using traditional and artifi cial intelligence approaches: discriminant analysis, logistic regression, neural networks and classifi cation and regression trees. Experimental studies using real world data sets have demonstrated that the classifi cation and regression trees and neural networks outperform the traditional credit scoring models in terms of predictive accuracy and type II errors.

show abstract

Section: Artifi Cial Intelligence Techniquesmentioning

confidence: 99%

A Comparison of Data Mining Techniques for Credit Scoring in Banking: A Managerial Perspective

İnce

Aktan

2009

Journal of Business Economics and Management

View full text Add to dashboard Cite

show abstract

“…Two adaptive neural networks were proposed and tested by Yang et al (2005) using simulated and measured data for building energy consumption prediction. Tso and Yau (2007) compared regression analysis, decision tree and neural network for the prediction of household electricity energy consumption in Hong Kong based on the electricity consumption survey data and obtained comparable results.…”

Section: Introductionmentioning

confidence: 89%

“…Specifically, it generally predicts the new events via the critical learning and training process on other existing events, especially useful in the situation of unknown relationship between the inputs and outputs (Tso, Yau 2007). A typical information processing unit in neural network is shown in Figure 2.…”

Section: Neural Network Modelmentioning

confidence: 99%

See 1 more Smart Citation

Lifecycle Energy Consumption Prediction of Residential Buildings by Incorporating Longitudinal Uncertainties

Wang

Shen

2014

Journal of Civil Engineering and Management

View full text Add to dashboard Cite

Abstract. Accurate prediction of buildings' lifecycle energy consumption is a critical part in lifecycle assessment of residential buildings. Longitudinal variations in building conditions, weather conditions and building's service life can cause significant deviation of the prediction from the real lifecycle energy consumption. The objective is to improve the accuracy of lifecycle energy consumption prediction by properly modelling the longitudinal variations in residential energy consumption model using Markov chain based stochastic approach. A stochastic Markov model considering longitudinal uncertainties in building condition, degree days, and service life is developed: 1) Building's service life is estimated through Markov deterioration curve derived from actual building condition data; 2) Neural Network is used to project periodic energy consumption distribution for each joint energy state of building condition and temperature state; 3) Lifecycle energy consumption is aggregated based on Markov process and the state probability. A case study on predicting lifecycle energy consumption of a residential building is presented using the proposed model and the result is compared to that of a traditional deterministic model and three years' measured annual energy consumptions. It shows that the former model generates much narrower distribution than the latter model when compared to the measured data, which indicates improved result.

show abstract

“…Depending on the availability and detail of data, each one has advantages and limitations: 1) dynamic models are based on solving mathematical equations derived from physical laws and require high detailed information of building materials; 2) bill-based methods requires less information but offer short options for innovation; 3) statistical methods do not require detailed information from the building and rely on training data to extract system function. For building energy forecasting, multiple regression, artificial neural networks (ANNs) and decision trees are the most used statistical techniques [7].…”

Section: Introductionmentioning

confidence: 99%

Energy consumption modeling by machine learning from daily activity metering in a hospital

Ruiz

Pacheco-Torres

Casillas

2017

2017 22nd IEEE International Conference on Emerging Technologies and Factory Automation (ETFA)

View full text Add to dashboard Cite

Hospitals are large buildings that consume a great amount of energy mostly due to their continuous energy consumption needs, high consumer medical equipment, and special requirements of thermal and air conditions. Reliable dynamic simulation is a chimera because of the complex design and behavior of these buildings. Therefore, monitoring-based methods arise as a plausible alternative. Its main drawback, however, is the lack of enough data to generate statistically robust models.The paper faces this problem thanks to the helpful contribution of a collaborative hospital which was able to generate daily data of electrical energy consumption for a period of six years. Besides, thirteen variables that summarize the daily activity of the hospital are also included. The results show how machine learning techniques generate models that accurately predict the electrical energy consumption based on weather conditions and activity measurements.The obtained results are useful for the design of more specific energy saving strategies, a more efficient economic investment for energy retrofitting of existing buildings and a better management of economic energy cost in large-scale buildings.

show abstract

Predicting electricity energy consumption: A comparison of regression analysis, decision tree and neural networks

Cited by 814 publications

References 11 publications

A Comparison of Data Mining Techniques for Credit Scoring in Banking: A Managerial Perspective

A Comparison of Data Mining Techniques for Credit Scoring in Banking: A Managerial Perspective

Lifecycle Energy Consumption Prediction of Residential Buildings by Incorporating Longitudinal Uncertainties

Energy consumption modeling by machine learning from daily activity metering in a hospital

Contact Info

Product

Resources

About