Gene Expression Programming (GEP) is a new adaptive algorithm that creates computer programs by using a genotype/phenotype genetic algorithm. GEP uses linear chromosomes composed of genes, which are structured into a head and a tail. These chromosomes function as genomes and are modified through mutation, transposition, gene transposition, gene recombination, and one- and two-point recombination. The chromosomes encode expression trees, which are selected based on fitness. GEP allows for high efficiency due to the distinct functions of the genome and expression tree. The algorithm is tested on various problems, including symbolic regression, sequence induction, block stacking, cellular automata rules for density classification, and boolean concept learning. GEP is a genetic algorithm that uses populations of individuals, selects them based on fitness, and introduces genetic variation through operators. Unlike genetic algorithms (GAs) and genetic programming (GP), GEP individuals are encoded as linear strings that are expressed as nonlinear entities (expression trees). This allows for a clear genotype/phenotype relationship, where modifications to the genome always result in valid expression trees. GEP chromosomes are composed of genes with fixed length, and each gene can code for expression trees of different sizes and shapes. GEP genes have a head and a tail, with the head containing functions and terminals, and the tail containing only terminals. The length of the head and tail is determined by the number of arguments of the function with the most arguments. GEP chromosomes can consist of multiple genes, each coding for a sub-expression tree. These sub-ETs can interact to form more complex expression trees. GEP uses various genetic operators, including replication, mutation, transposition, insertion sequence elements, root transposition, gene transposition, and recombination, to introduce genetic diversity and evolve solutions. GEP is used to solve a variety of problems, including symbolic regression, planning, boolean concept learning, and cellular automata rules. The algorithm is tested on complex problems, such as the density-classification task, where GEP surpasses GP by more than four orders of magnitude. GEP is capable of evolving complex solutions through the modular construction of hierarchical structures, where each gene codes for a small building block. The success of GEP depends on the design of fitness functions, which determine the goal of the evolutionary process. GEP is also effective in finding optimal solutions through the use of appropriate linking functions and genetic operators. The algorithm is versatile and efficient, capable of adapting and evolving solutions to complex problems.Gene Expression Programming (GEP) is a new adaptive algorithm that creates computer programs by using a genotype/phenotype genetic algorithm. GEP uses linear chromosomes composed of genes, which are structured into a head and a tail. These chromosomes function as genomes and are modified through mutation, transposition, gene transposition, gene recombination, and one- and two-point recombination. The chromosomes encode expression trees, which are selected based on fitness. GEP allows for high efficiency due to the distinct functions of the genome and expression tree. The algorithm is tested on various problems, including symbolic regression, sequence induction, block stacking, cellular automata rules for density classification, and boolean concept learning. GEP is a genetic algorithm that uses populations of individuals, selects them based on fitness, and introduces genetic variation through operators. Unlike genetic algorithms (GAs) and genetic programming (GP), GEP individuals are encoded as linear strings that are expressed as nonlinear entities (expression trees). This allows for a clear genotype/phenotype relationship, where modifications to the genome always result in valid expression trees. GEP chromosomes are composed of genes with fixed length, and each gene can code for expression trees of different sizes and shapes. GEP genes have a head and a tail, with the head containing functions and terminals, and the tail containing only terminals. The length of the head and tail is determined by the number of arguments of the function with the most arguments. GEP chromosomes can consist of multiple genes, each coding for a sub-expression tree. These sub-ETs can interact to form more complex expression trees. GEP uses various genetic operators, including replication, mutation, transposition, insertion sequence elements, root transposition, gene transposition, and recombination, to introduce genetic diversity and evolve solutions. GEP is used to solve a variety of problems, including symbolic regression, planning, boolean concept learning, and cellular automata rules. The algorithm is tested on complex problems, such as the density-classification task, where GEP surpasses GP by more than four orders of magnitude. GEP is capable of evolving complex solutions through the modular construction of hierarchical structures, where each gene codes for a small building block. The success of GEP depends on the design of fitness functions, which determine the goal of the evolutionary process. GEP is also effective in finding optimal solutions through the use of appropriate linking functions and genetic operators. The algorithm is versatile and efficient, capable of adapting and evolving solutions to complex problems.