JP Journal of Heat and Mass Transfer
Special Volume, Issue II, Advances in Mechanical System and ICT-convergence, Pages 151 - 156
(July 2018) http://dx.doi.org/10.17654/HMSI218151 |
|
WEB CACHE OPTIMIZATION WITH BAYESIAN AND MAXIMUM LIKELIHOOD ESTIMATION
Prapai Sridama
|
Abstract: The objective of this research increases efficiency of web cache memory. The web cache optimization with Bayesian and maximum likelihood estimation (WCO-BMLE) simulation is investigated for decision making about web objects replacement based on web usage and dynamic decision. WCO-BMLE model used web objects data more than 30 weeks for prediction of web usage probability. In addition, many statistics mathematics theories are used within WCO-BMLE model as follows: interpolation with cubic spline for curve fitting lines, an agent finding of each web usage objects with expected value algorithm, first order condition (FOC) using for trend study of web usage in the past. However, FOC is a technique to decide the replacement. WCO-BMLE model is a replacement technique that solves replacement problem better than the least recently used (LRU) technique. Though, dynamic programming with Bellman equation is last technique when others technique cannot solve replacement problems. In addition, Bayesian algorithm and maximum likelihood estimation are raised when web usage from clients is different from in the past. |
Keywords and phrases: computing, cubic spline, Bellman equation, Bayesian. |
|
Number of Downloads: 376 | Number of Views: 3042 |
|