Convert Probability Network to Artificial Neural Network based on Position, Time and Speed of Events

9th Annual Basic Science International Conference 2019

自然科学（总论）

Cholissodin, Imam^1 ; Marji^1

Faculty of Computer Science, Brawijaya University, Malang, Indonesia^1

关键词: Bayes theorem;    Concept comparison;    First order logic;    Probability functions;    Propositional logic;    Set events;    Time features;   

Others  :  https://iopscience.iop.org/article/10.1088/1757-899X/546/5/052018/pdf
DOI  :  10.1088/1757-899X/546/5/052018

学科分类：自然科学（综合）

来源: IOP

【 摘 要 】

Building a probability function to get the best answer between events in many review problems is a very complex one. In this study, several events were given as a case study to find a solution, after which an analysis will be carried out when and where the event contained position and time features. So that the distance factor, travel time factor and speed factor for each event can be calculated. Then a concept comparison between Probability Network from the Bayes Theorem versus Artificial Neural Network (ANN) was made, accompanied by a review of the Propositional Logic (PL) or First Order Logic (FOL) concept compared to using the Uncertainty concept in calculating the probability of an eventin a pseudo event that has often been ignored because it is considered to have no influence and the probability is small whereas it can be the opposite. In fact, sometimes, the pseudo event is the main factor determining the real event and the probability is much greater than expected. In addition, we also give an Engineering Events Theorem that can intentionally set event-by-event syntheses to increase the probability of an event which is initially smaller than the probability of another event. After being given an injection/dummy/synthesis event, however, then the initially small probability of the event becomes very high and the time of occurrence can be accelerated or decelerated, or even minimized so that it does not occur.