/*****************************************************************
- READ ME:
- Evolutionary Algorithms
- An Implementation of Genetic Algorithms
- and Population-Based Incremental Learning
- By Tyler Higgins, Marissa Rosenthal, and Ruben Martinez
- Included Files:
- SATFileReader.java
- EvolAlg.java
- Required Arguments:
- For Genetic Algorithms:
- [1] Name of File
- i.e. file.cnf
- [2] Number of Individuals (Population)
-
i.e. 100 - [3] Selection Method
-
[a] Tournament select ( ts ) -
[b] Rank select ( rs ) -
[c] Boltzmann ( bs ) - [4] Crossover Method
-
[a] One-Point Crossover ( 1c ) -
[b] Uniform Crossover ( uc ) - [5] Crossover Probability
-
i.e. 0.7 - [6] Mutation Probability
-
i.e. 0.02 - [7] Number of Generations
-
i.e. 100 - [8] Algorithm
-
[a] Genetic Algorithm ( g ) - Example: java EvolAlg file.cnf 100 ts 1c 0.7 0.01 100 g
- For Population-Based Incremental Learning:
- [1] Name of File
-
i.e. file.cnf - [2] Number of Individuals (Population)
-
i.e. 100 - [3] Positive Learning Rate
- i.e. 0.1
- [4] Negative Learning Rate
-
i.e. 0.075 - [5] Mutation Probability
-
i.e. 0.02 - [6] Mutation Amount
-
i.e. 0.05 - [7] Number of Generations
- i.e. 100
- [8] Algorithm
- [a] PBIL ( p )
- Example: java EvolAlg file.cnf 100 0.1 0.075 0.02 0.05 100 p
- Instructions:
- Our program takes in a file containing formatted MAXSAT
- problems. Depending on the arguments, it will either run
- a genetic algorithm, or run a population-based incremental
- learning algorithm for a specified number of generations,
- and outputs the best individual created by the algorithm,
- the generation it was found, the number of clauses it
- satisfies, the percent satisfied clauses, as well as some
- information about the input file such as name, number of
- variables, number of clauses, and the maximum number of
- variables in a clause.
- When running a genetic algorithm, the user can choose among
- tournament selection, rank selection, or Boltzmann selection.
- They can also choose between one-point crossover and uniform
- crossover, as well as select custom values for population
- size, number of generations, and crossover and mutation
- probabilities.
- When running in PBIL mode, the user can also choose a custom
- population size, number of generations, and mutation amount.
- They can also specify positive and negative learning rates,
- as well as, mutation amount.
*****************************************************************/