This course provides clear explanations and hands-on examples that make complex AI concepts easy to understand.
Course Participant
Student
What you will learn?
In this foundational section, students will gain a clear understanding of the basics of optimization, including its definitions, importance, and core principles. We explore why optimization is essential across various fields, how it helps improve efficiency, reduce costs, and enhance performance, and what makes it a critical part of intelligent systems. Learners will be introduced to key concepts such as objective functions, critical points, local and global optima, convergence, and the balance between exploration and exploitation. The section also covers some metaheuristic algorithms, along with intuitive examples of how they work. Finally, students will learn about benchmark test functions used to evaluate optimization performance, setting the stage for deeper algorithmic exploration in future sections.
In this lecture, you’ll simulate protein folding using the Differential Evolution (DE) algorithm. You’ll learn how to represent proteins as amino acid chains in 3D space and optimize their positions to minimize total energy. By the end, you’ll be able to predict stable protein structures computationally and understand how misfolding can lead to diseases like Alzheimer’s and Parkinson’s.
In this section, you’ll explore how the Firefly Algorithm (FA) can be used to solve a physics-inspired optimization problem based on space-time warping. You’ll learn how concepts such as curvature, bending effort, warp fields, and geodesics can be translated into a mathematical objective function that guides fireflies to find the most energy-efficient path through a distorted space. By simulating this warped environment, each firefly represents a potential solution that evolves over time using brightness-based attraction, distance decay, and controlled randomness. The algorithm works to minimize a total energy cost that includes warped distance, curvature penalties, traversal effort, and warp-field maintenance, ultimately finding a smooth, optimal path between two points. This section combines physical intuition with computational intelligence, showing how bio-inspired algorithms can solve problems modeled after the curvature and dynamics of space-time.
By the end of this section, students will understand how living organisms adapt to diverse exoplanetary environments. They will be able to explain how planetary factors like gravity, radiation, temperature, and atmosphere affect survival traits, and how these traits interact to determine overall fitness. Students will also learn to interpret and analyze adaptation outcomes, comparing evolved traits with environmental requirements to identify successful or failed survival strategies.
This section explains how optimization algorithms can be applied to a real engineering problem: automatically shaping antennas to achieve high efficiency, minimal material use, and smooth geometry. Students learn how PSO mimics the collective intelligence of swarms to explore complex 3D design spaces that are difficult to optimize analytically. The section connects mathematical formulation with real implementation, showing how total length and bending smoothness can be combined into a single objective function and minimized iteratively through PSO. After completing this section, students will understand how to formulate and solve antenna design problems using metaheuristic methods, interpret convergence behavior, visualize optimized 3D geometries, and appreciate the power of swarm-based algorithms in electromagnetic and structural design optimization.
About the course
Optimization is at the heart of modern science, engineering, and artificial intelligence. From designing antennas for space exploration to training cutting-edge AI models, optimization algorithms provide powerful tools to solve problems once thought impossible. Yet, most courses only scratch the surface with formulas and theory; this course is different.
In Optimization Algorithms for Real-World Problems, you will learn not just the basics of optimization but also how to apply advanced algorithms to practical scenarios. We’ll explore Genetic Algorithms, Particle Swarm Optimization, Differential Evolution, and Firefly Algorithms. Each concept is explained with clear examples, simulations, and case studies drawn from diverse fields such as biology, aerospace, computer science, data science, robotics, and engineering.
This is an intermediate course. Whether you are a Bachelor's student beginning your research journey, a Master’s or PhD student deepening your expertise, or a professional researcher or engineer looking to enhance your toolkit, this course is designed for you. No matter your level, you’ll gain the ability to understand, formulate, and solve your own optimization problems.
By the end of this course, you will have both the theoretical foundation and the practical skills needed to apply optimization to real-world challenges, bridging the gap between equations, simulations, and truly impactful results.
About the teacher
Seyed Muhammad Hossein Mousavi
Senior Researcher in A.I. Emotion Recognition, Depth Image Processing, Optimization, Virtual...
Seyed Muhammad Hossein Mousavi is an advanced AI researcher and innovator with a strong focus on optimization, multimodal learning, and emotion recognition. With over 50 peer-reviewed publications and a Google Scholar H-index of 12, he is ranked among the top 6% of active researchers in AI globally, according to ResearchGate metrics.
Mr. Mousavi is known for his pioneering work in emotion recognition from various modalities, synthetic data generation, and meta-heuristic optimization algorithms. His research bridges theory and real-world application, from emotion recognition using body motion and physiological signals to VR-based stimulus systems, depth estimation, pose optimization, and hardware-integrated multimodal AI systems.
As the founder of Cyrus Intelligence Research, he has led several independent projects combining Kinect motion capture, Arduino-based sensors, VR protocols, and AI pipelines involving techniques like PSO, GA, GCNs, ANFIS, and GANs. He has also contributed to open-source research and published reproducible codebases on GitHub.
Mr. Mousavi’s work spans multimodal fusion, emotion-driven synthetic data, graph neural networks, and real-time optimization for health and affective computing, with datasets and frameworks designed for academic and industrial AI use.
In addition to research, he is passionate about scientific transparency, hands-on experimentation, and democratizing access to AI knowledge through tutorials, pipelines, and public datasets.
