Twitter | Search | |
Search Refresh
Graham Kendall May 22
"As we increase the number of pieces in the training data to 250 and 500, we see an increase in performance in the evolved heuristic"
Reply Retweet Like
Graham Kendall May 22
"Our results showed that the simulated wheel was fair and balanced in this respect"
Reply Retweet Like
Graham Kendall May 22
"We explain the 'best fit' algorithm, examine a graph of ?best fit? (figure 3), and explain intuitively why it is a good heuristic"
Reply Retweet Like
Graham Kendall May 16
"The custom bias enables us to target specific numbers, which means that certain numbers are more likely to appear than others"
Reply Retweet Like
Graham Kendall May 16
"The decision of which problem-specific structures, operators, and parameter values to be included (or excluded) in a given meta-heuristic algorithm has an impact on algorithm performance"
Reply Retweet Like
Graham Kendall May 22
" and has many real-life practical applications - paper, metal, glass or other sheet material cutting, basic pallet loading, etc., and several theoretical ideas such as multi-processor scheduling."
Reply Retweet Like
Graham Kendall May 22
"If you were to drill out the number one, this means that the number six is more likely to appear, as six is on the opposite face to one"
Reply Retweet Like
Graham Kendall May 17
"As Table 1 demonstrates, even after 1.5 billion spins there is still some deviation from theoretical expectations"
Reply Retweet Like
Graham Kendall May 16
"Now we can make the additional claim that we can apply our heuristics to problems of much larger size, without deterioration of solution quality"
Reply Retweet Like
Graham Kendall May 16
"Commonly used soft constraints include spreading conflicting exams as evenly as possible, or not in x consecutive timeslots or days"
Reply Retweet Like
Graham Kendall May 16
"In addition, the proposed algorithm adheres to all the hard constraints which the current methodology fails to do"
Reply Retweet Like
Graham Kendall May 15
"A further contribution of this work is the formulation of the UMP examination timetabling problem as a mathematical model"
Reply Retweet Like
Graham Kendall May 16
"As educational AI develops, students will be able to study where they want, when they want and using whatever platform they want"
Reply Retweet Like
Graham Kendall May 17
"The problem is complicated by the fact that the chief invigilator position can only be assigned to academic staff and staff are not allowed to invigilate their own papers."
Reply Retweet Like
Graham Kendall May 17
"... draws inspiration from the equivalent two-dimensional problem, whilst adding a number of substantial changes and improvements in order to make it applicable to the three dimensional problem."
Reply Retweet Like
Graham Kendall May 16
"Over the festive period every team plays two fixtures. Whilst scheduling these two sets of fixtures the overriding aim is to minimise the total distance that has to be travelled by the supporters."
Reply Retweet Like
Graham Kendall May 16
"They also tend to actively promote themselves through email campaigns, have nonexistent peer review and have affordable publication fees when compared to legitimate open access journals"
Reply Retweet Like
Graham Kendall May 16
"We note that there is a gap in terms of the examination timetabling datasets from the literature and many of the requirements faced by many institutions"
Reply Retweet Like
Graham Kendall May 18
"After empirical testing over a range of parameter rates, we use 0.6 for crossover rate, 0.1 for mutation rate, a population size of 30, 200 generations"
Reply Retweet Like
Graham Kendall May 16
"Three new operators: best-best crossover, removing-worst mutation, and inserting-good mutation help to dynamically insert or remove heuristics"
Reply Retweet Like