Süni intellektTopics

In the subject, the origin of artificial intelligence, historical development stages and fundamental qualities of intelligence are studied. students analyze the turing test and chinese room philosophical arguments, as well as developmental schools, to understand the essence of ai. in the course of the lesson, concepts of different approaches to ai, functional types of ai, the concept of augmented intelligence and global ethical principles in this field are examined.

The subject studies expert systems (es), which are one of the most successful practical applications of artificial intelligence, simulating the decision-making ability of a human expert. students analyze the characteristics of expert systems and examine their comparative differences with human experts, traditional software, and neural networks. in the course of the lesson, the stages of creation of ES and its architectural components — knowledge base, inference engine that performs logical inferences, and working memory that stores current data are studied in depth. also, the types of knowledge that are the basis of knowledge engineering, the differences between obvious and non-obvious knowledge, as well as the representation models of knowledge are analyzed.

In the topic, the internal architecture of expert systems (es), which is one of the fields of application of artificial intelligence, and the formation steps as a whole software are studied in depth. students analyze the three fundamental components that form the core of es—the knowledge base, which stores domain-specific rules and facts, the inference engine, which makes logical inferences based on existing knowledge, and the working memory structures, which process momentary data relevant to the current problem. also, the engineering stages of an expert system from design to implementation are examined in sequence.

In the subject, the basic concepts and methodology of statistics, which are the main pillars of data science and machine learning, are studied. students analyze the principles of collecting, arranging and analyzing quantitative indicators, as well as the fundamental differences between the main population, which represents the entire population of the research object, and the sample population, which is a small group selected from it according to certain rules. also, the types of variables used in statistical analyzes and their measurement scales are examined.

In the subject, the basic statistical indicators that determine the central tendency and distribution characteristics of the data set, as well as the degrees of dependence between the variables, are studied. students analyze the concepts of mean, mode, median, which determine the center of data, and dispersion and mean squared deviation, which measure how spread out the data are. in the second part of the lesson, the principles of covariance, which shows how two different variables change together, and coefficient of determination, which determines how much one variable explains the other, are examined.

The subject studies the field of machine learning, which is the most dynamic branch of artificial intelligence, which enables systems to learn based on data without programming. students analyze the historical evolution, basic terminology, and fundamental methods of machine learning. the main part of the lesson explores the architectural stages that make up the end-to-end life cycle of a machine learning project — data acquisition, data preprocessing, modeling, execution/evaluation, and the principles of releasing the finished model to a production environment.

The topic explores the most widely applied field of machine learning in industry — the concept of supervised learning. students analyze a "tutored learning" mechanism in which input data and corresponding correct output responses are pre-fed to the model. in the course of the lesson, the logic of regression and classification, which are the two main directions of controlled learning, as well as the fundamental methods used in these directions are examined.

Classification and regression tasks, two main functional areas of supervised learning, fundamental differences between them and application scenarios in industry are studied in the topic. students compare the logic of classification, where the target variable is discrete categories, to regression, where the target variable is continuous, real numerical values. the second part of the lesson examines the basic regression algorithms used to model linear, multivariate and non-linear dependencies.

The topic explores the concept of unsupervised learning, a branch of machine learning that works with unlabeled raw data. students analyze ways to discover hidden structures, internal regularities, and similarities in data where output responses are not provided in advance. the main part of the lesson examines the clustering method, which divides the dataset into subgroups based on similar characteristics, its working principles, customer segmentation in the industry, and anomaly detection.

In the subject, machine learning and regression analysis, one of the most fundamental forecasting methods of statistics, are studied. students analyze the mathematical relationships between dependent and independent variables, the mechanism of continuous numerical value prediction, and the classification of regression. the main part of the lesson examines the working principles, equations and evaluation of simple linear regression based on only one independent variable and multivariate linear regression models that measure the effect of multiple factors on the target variable.

The topic explores the concept of fuzzy logic, which transcends the boundaries of classical binary logic and allows modeling real-world uncertainties. students comparatively analyze the principles of fuzzy logic, which accepts an infinite number of continuous degrees of association between zero and one, with traditional logic based only on true or false values. in the main part of the lesson, the concept of fuzzy set, membership functions and the rules of applying fundamental mathematical operations on these sets are examined.

The subject studies the biological and mathematical connection between the working mechanism of the human brain and neural networks, the most powerful branch of artificial intelligence. students analyze the structure of the biological neuron, which is the basis of the central nervous system, its parts, and the principles of signal transmission. the main part of the lesson examines the first artificial neuron models, which are engineering adaptations of this structure — the mcculloch-pitts and perceptron models, as well as the generalized mathematical components of an artificial neuron. finally, the network models and types formed by these single neurons are analyzed.

The subject studies the fastest growing direction of artificial intelligence and machine learning, based on multilayer artificial neural networks — deep learning. students analyze deep learning's fundamental concepts, specific terminology, historical evolution, and how it differs from other classes in machine learning. the main part of the lesson examines the main deep learning methods and the structure of different network architectures used to process complex, non-linear data and large data sets.

In the topic, the main types of models of deep learning applied in the industry and their different characteristics are studied. students compare the fundamental differences between traditional machine learning and deep learning—feature extraction process, data volume dependency, computing resources, and performance metrics. the main part of the lesson examines the real application areas that shape the modern world, such as computer vision, natural language processing, speech recognition and autonomous systems.

The subject studies genetic algorithms inspired by the natural selection and evolution processes in nature and applied to solve complex search and optimization problems in an artificial intelligence environment. students analyze fundamental concepts such as population formation, chromosomal encoding of individuals, and a fitness function that measures the quality of candidate solutions. in the main part of the lesson, the mechanisms of selecting the best individuals for the next generation, crossover to create new solutions, working principles of mutation operators to preserve genetic diversity and avoid local minima, as well as elitism strategy are practically investigated.