INVESTIGATION OF THE INFLUENCE OF ANNEALING MODES OF THE GAAS(111) SURFACE ON THE CHARACTERISTICS OF NANOHOLES FORMED BY FOCUSED ION BEAMS AT VARIOUS EXPOSURE TIMES

The popularity of container systems attracts the attention of many researchers in the field of informationtechnology. Containerization technology allows to reduce the cost of computing resources whendeploying and supporting complex infrastructure solutions. Ensuring the security of container systems andcontainerization in general, as well as the use of smart attacks based on artificial intelligence by malefactors,is a serious problem on the way to the safe and stable operation of container systems. This articleproposes an approach for detecting not only previously unknown individual anomalous processes, but alsoanomalous process sequences in container systems. The proposed approach and its implementation basedon the Docker platform are based on tracing system calls, constructing histograms of running processes,and using the AE-LSTM neural network. The process of constructing histograms is based on accounting ofthe number of executed system calls for each individual process. This solution provides the ability not onlyto accurately identify any process in the system, but also to effectively detect anomalous process sequenceswith a high degree of accuracy. The generated sequences are used as input data for the neural network.After completing the training process, the neural network acquires the ability to detect anomalous sequencesby comparing a given threshold of reconstruction error with the actual error level of the inputdata vector. When the neural network encounters a new input data vector, it calculates the reconstructionerror level - the difference between the expected and actual value. If this error exceeds a predetermined threshold, the system signals the presence of an anomaly in the sequence. Experiments show that the proposedapproach demonstrates high accuracy in detecting anomalous processes with a low level of falsepositive detection results. Such results confirm the effectiveness of the proposed approach. Also, the computationalcosts of training the neural network model are quite low. This allows using less powerful hardwarewithout significant performance losses. Such a solution can be trained and implemented in a newinfrastructure in a fairly short time