Developing Predictive Models using Typical Machine Learning and Computational Techniques
Journal Title: Annals. Computer Science Series - Year 2018, Vol 16, Issue 2
Abstract
This study investigates the accuracy of developing predictive models using machine learning techniques. The machine learning techniques considered in this study include artificial neural network (ANN) and Kalman filter adaptation algorithm. Predictive values are computed based on these techniques. These techniques are tested on daily electricity consumption data and are computed using ANN technique and Kalman filter adaptation algorithm. The accuracy of the predicted values of these techniques are investigated using statistical parameters. This research identified Kalman technique as more accurate in making predictions than ANN technique.
Authors and Affiliations
Patrick A. OZOH, Shapiee ABD-RAHMAN, Morufu Oyedunsi OLAYIWOLA
Keywords
Related Articles
- EP ID EP550031
- DOI -
- Views 93
- Downloads 0
Computerized Health Information System
Today, medical doctors are dealing with more complex diagnostic methodologies, and with increasing volumes of medical information related to delivery of health care in various medical centres and institutions around the...
Iterative Methods for Systems’ Solving - a C# approach
This work wishes to support various mathematical issues concerning the iterative methods with the help of new programming languages. We consider a way to show how problems in math have an answer by using different academ...
The certification of the EHR systems according to EuroRec Seal
As in any field, in the e-health domain it is very important that the designing of a computer system should start by establishing a complete set of requirements for the architecture of the future EHR system. The present...
E-learning and use of computer in forensic field
In the Romanian penitentiary establishments and those of other European countries one talks about the formal and informal education. Since the correspondence education system, important steps have been made towards the e...
Proposal of a Qualification and Selection of a Software for Analysis and Monitoring of Search Engines
In order to measure website visibility in search engines, there are softwares for analytics and referencing follow-up. They permit to quantify website's efficacity of referencing and optimize its positionning in search e...