Intelligent algorithms that allow systems to learn and reason, especially those used to analyze images and video and computer vision, and in a more general sense – artificial intelligence (AI). For several decades, the subject of intensive research, both in theory and in practice. This course is intended for students preparing a master’s degree in artificial intelligence and intelligent systems, especially for those who want to complete the final project, as well as for everyone who wants to learn about creating models designed for AI, computer vision and image and video analysis ; and even people working on information support and decision-making systems and data mining models. We hope that this book will be an effective contribution, offering students a collection of simplified, understandable ideas about writing intelligent algorithms. In addition, presenting these concepts in a sober form, he easily leads the reader to understand them, thereby providing him with the opportunity to specifically control the automatic construction of the model.
A brief history in the field of artificial learning, also commonly referred to as machine learning (AA), prompts us to talk about the three great eras of computer, more precisely, at the very beginning of computing, their evolution in time and, finally, in the world today and tomorrow. At present, we can note, and this is only the point of view, that the evolution of computer science was done mainly along two axes: Increasing the ability to accumulate information and its dissemination in areas such as Data Mining, data storage, networks and web services, not forgetting at the same time, their basic applications on smartphones, on the one hand, and acquiring the intellectual capabilities of computer systems, in particular, the areas related to artificial intelligence, which are most affected by this technological progress and, in particular, including games, speech, computer vision, etc.
In this subject, we are interested in theories, algorithms, and applications related to a specific aspect of artificial intelligence (AI): the faculty of education. From this concept of learning, it becomes easy to see that the paradigm of classical programming has evolved, therefore: Before that, programming consisted of prescriptive logic for the execution (task). Now we plan to create an intelligent, in other words, program for the intellectual fulfillment of new tasks, which are considered to require reasoning and judgment. Thanks to this programming mechanism, we will be able to equip the program with learning abilities.
The discipline of machine learning (AA) has a rich theoretical foundation. Now we know how to answer questions such as: – What teaching methods are most effective for a particular type of problem? – How many training examples do you need to provide for the curriculum to be sure that you are learning with the given effectiveness?
Among the main characteristics and abilities adopted in training models, we note: training, generalization, adaptation, improvement, comprehensibility and forecasting.
In practice, the word learning is often associated with: learning. Thus, in cognitive science, learning is defined as the ability to improve productivity as activities progress. For example, a chess player: assimilates (through experience, coaches) and reasons (this gives him some intelligence or the ability to reason so that he can progress) – this is the case of an intelligent algorithm.
This can be considered as a model model (algorithm or system) for correcting its behavior or changing its reaction (for example, forecasting) in connection with new situations. For the tasks of perception in artificial vision, we accumulate good and bad experiences, and from them we can change the rules to better perform the task, this is the phenomenon of adaptation or improvement.
Why do experts discuss digital aids? From the point of view of data mining, it is simply a matter of controlling the intelligibility (clarity) of the resulting model. At present, the measurement of intelligibility is limited to verifying that the knowledge gained is clear and that the results are expressed in the user’s language, and the size of the models is not excessive.
Additionally. Having professional experience over several years, AA determines the security policy of the information system and ensures its implementation. It plays the role of counseling, assistance, information, training and alert management. Depending on the size of the facility, he / she plays an operational role in implementing the security policy or leads a team of technical experts and consultants. He / she offers the competent authority an IP security policy and ensures its implementation. He / she can intervene in SSI issues at all or part of the IT and telecommunication systems of his unit, both at the technical and organizational level.