Simulating Natural Selection with Deep Learning and Genetic Algorithm
Piyush Pal, Shashank Mishra, Mudit Pratap Singh, Shivani Trivedi · 2023
Bridging the realms of biology and artificial intelligence, our study introduces a comprehensive simulation. Within this virtual ecosystem, creatures undergo generational evolution, continually adapting to their dynamic environment. Central to this simulation is an innovative neural network module that equips these creatures with real-time predator sensing and decision-making capabilities. Our approach encompasses diverse creature generation, rigorous fitness evaluation, genetic algorithm-driven evolution, and intricate neural network design. Through extensive experimentation, our findings unveil the intricate dynamics of adaptation and decision-making, illuminating the pivotal role of sensory perception in the evolutionary process. This work significantly advances our comprehension of natural selection and AI-driven decision-making within controlled ecosystems. It not only enriches interdisciplinary insights but also serves as a catalyst for future research endeavors.