Assessing the predictive performance of artifIcial neural network-based classifiers based on different data preprocessing methods, distributions and training mechanisms

An accurate demand forecasting is essential for planning the electric dispatch in power system, contributing financially to electricity companies and helping in the security and continuity of electricity supply. In addition, it is evident that the distributed energy resource integration in the electric power system has been increasing recently, mostly from the photovoltaic generation, resulting in a gradual change of the load curve profile. Therefore, the 24 hours ahead prediction of the electrical demand of Campina Grande, Brazil, was realized from artificial neural network with a focus on the data preprocessing. Thus, the time series variations, such as hourly, diary and seasonal, were reduced in order to obtain a better demand prediction. Finally, it was compared the results between the forecasting with the preprocessing application and the prediction without the preprocessing stage. Based on the results, the first methodology presented lower mean absolute percentage error with 7.95% against 10.33% of the second one.

Download Full-text

ANNETT-O: an ontology for describing artificial neural network evaluation, topology and training

International Journal of Metadata Semantics and Ontologies ◽

10.1504/ijmso.2019.099833 ◽

2019 ◽

Vol 13 (3) ◽

pp. 179

Author(s):

Iraklis A. Klampanos ◽

Athanasios Davvetas ◽

Antonis Koukourikos ◽

Vangelis Karkaletsis

Keyword(s):

Neural Network ◽

Artificial Neural Network ◽

Artificial Neural ◽

Network Evaluation ◽

And Training

Download Full-text

Artificial Neural Network Based Step-Length Prediction Using Ultrasonic Sensors from Simulation to Implementation in Shoe-Type Measurement Device

Journal of Advanced Computational Intelligence and Intelligent Informatics ◽

10.20965/jaciii.2017.p0321 ◽

2017 ◽

Vol 21 (2) ◽

pp. 321-329 ◽

Cited By ~ 2

Author(s):

Romy Budhi Widodo ◽

◽

Chikamune Wada ◽

Keyword(s):

Neural Network ◽

Artificial Neural Network ◽

Physical Therapist ◽

Predictive Performance ◽

Step Length ◽

Ultrasonic Sensors ◽

Measurement Device ◽

Gait Parameters ◽

Artificial Neural ◽

Artificial Neural Network Ann

Step-length measurement as a spatial gait parameter is useful for the physician and physical therapist for determining the patient’s gait condition. We hypothesized that this could be determined using ultrasonic sensors mounted on a shoe-type measurement device. For that purpose, we have developed a shoe-type measurement device to measure gait parameters. Our system was found to effectively measure step-length and pressure distribution. However, we found that the presence of shoes leads to perishable and fragile conditions for the sensors. Therefore, we redesigned the number, angle, and range of the ultrasonic sensors mounted on the shoes in order to clarify and improve the step-length prediction. This paper discusses the improvement of a shoe-type measurement device from the implementation with real shoes and the step-length prediction using an artificial neural network (ANN). The results of the experiment show that the number, angle, and positioning of ultrasonic sensors affect their ability to capture the human step region, that is, 50×70 cm under the experimental condition of foot progression angle up to 30 degrees. The results of the predictive performance of step-length using the proposed ANN architecture demonstrate an improvement.

Download Full-text