Skip to content

akmaralAW/Multidimensional_Knapsack

Repository files navigation

Multidimensional_Knapsack

Task: Design a dynamic programming algorithm that takes a tuple {(damage, red, blue), (damage, red, blue), ...} as an input. Play cards so the total amount of damage is maximized, while having limited red, blue cards.

Solution: 0/1 Knapsack based solution. Table has nxredxblue dimension. T[i, j, k] - "maximum damage value" of j-th red, k-th blue, considering all values from 0 to i-th participant.

About

0/1 Knapsack based solution. More details will be soon.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published