PBS Colors?
#11
Line Holder
Thread Starter
Joined APC: Aug 2016
Posts: 44
PBS Core Algorithms
PBS Core Algorithms
[Greedy]
This algorithm uses a "First Come First Serve" concept. For each crewmember in a group of the same bid package, starting with the most senior, it awards the best legal line based on his/her preferences without a conflict with his/her carry-in and pre-defined activities, then removes the awarded pairings from those available. In the same way it awards the best legal line for the next crewmember. This process is repeated until the most junior crewmember has been allocated a line.
The best line is defined by one with the most prioritized preferences or the most points (if it is a weighted/pointed system) of the preferred pairings and line characteristics. (In the case of more than one equal important best lines available, an arbitrary one will be awarded.)
Greedy Algorithm does not consider the preferences of junior crewmembers, it offers lowest overall satisfaction.
Using this algorithm alone will not control open time. Uncovered pairings will be left on any date throughout the bid month, especially during holiday periods as a result of most crewmembers preferring off on those days. Therefore, it requires more reserve crewmembers to cover these uncovered pairings.
In brief, the greedy algorithm cannot achieve high overall satisfaction and cannot control open time. It cannot create the line holders required. Unfortunately, it is the fundamental algorithm being used by almost every PBS on the market today.
[Greedy]
This algorithm uses a "First Come First Serve" concept. For each crewmember in a group of the same bid package, starting with the most senior, it awards the best legal line based on his/her preferences without a conflict with his/her carry-in and pre-defined activities, then removes the awarded pairings from those available. In the same way it awards the best legal line for the next crewmember. This process is repeated until the most junior crewmember has been allocated a line.
The best line is defined by one with the most prioritized preferences or the most points (if it is a weighted/pointed system) of the preferred pairings and line characteristics. (In the case of more than one equal important best lines available, an arbitrary one will be awarded.)
Greedy Algorithm does not consider the preferences of junior crewmembers, it offers lowest overall satisfaction.
Using this algorithm alone will not control open time. Uncovered pairings will be left on any date throughout the bid month, especially during holiday periods as a result of most crewmembers preferring off on those days. Therefore, it requires more reserve crewmembers to cover these uncovered pairings.
In brief, the greedy algorithm cannot achieve high overall satisfaction and cannot control open time. It cannot create the line holders required. Unfortunately, it is the fundamental algorithm being used by almost every PBS on the market today.
#12
Gets Weekends Off
Joined APC: May 2017
Posts: 2,145
"[Un-Stacking]
This algorithm is used in conjunction with greedy algorithm to control the open time (the amount of uncovered pairings at each calendar date).
"Stacking" happens during the process of junior awarding when the number of the pairings needed to be covered at a given date is equal to the number of available crewmembers. The "available crewmembers" means those junior crewmembers who have not been awarded yet and are legal to cover a pairing at that date. Each given date in a bid month will have a "stack" to control open time.
When that point is reached, every available junior crewmember has to cover a pairing at that date to resolve the open time issue regardless of his/her preferences. This impacts the satisfaction of junior crewmembers in order to achieve this company goal.
To improve, the core traces back to the awarded senior crewmembers and sees if it is possible to let a senior cover a pairing at that date. This approach is called "un-stacking", because it reduces the stack at the given date. It is only good if the pairing being replaced is something junior crewmembers prefer, and the replacement does not degrade the senior crewmember's preferences and there is no violation of legality.
Greedy algorithm has no ability to constantly monitor the open time, and it keeps awarding lines until it reaches the stacking point, then uses inefficient "un-stacking" to improve the satisfaction of junior crewmembers. The result is questionable."
This algorithm is used in conjunction with greedy algorithm to control the open time (the amount of uncovered pairings at each calendar date).
"Stacking" happens during the process of junior awarding when the number of the pairings needed to be covered at a given date is equal to the number of available crewmembers. The "available crewmembers" means those junior crewmembers who have not been awarded yet and are legal to cover a pairing at that date. Each given date in a bid month will have a "stack" to control open time.
When that point is reached, every available junior crewmember has to cover a pairing at that date to resolve the open time issue regardless of his/her preferences. This impacts the satisfaction of junior crewmembers in order to achieve this company goal.
To improve, the core traces back to the awarded senior crewmembers and sees if it is possible to let a senior cover a pairing at that date. This approach is called "un-stacking", because it reduces the stack at the given date. It is only good if the pairing being replaced is something junior crewmembers prefer, and the replacement does not degrade the senior crewmember's preferences and there is no violation of legality.
Greedy algorithm has no ability to constantly monitor the open time, and it keeps awarding lines until it reaches the stacking point, then uses inefficient "un-stacking" to improve the satisfaction of junior crewmembers. The result is questionable."
Thread
Thread Starter
Forum
Replies
Last Post