International Scientific Journals
of Scientific Technical Union of Mechanical Engineering "Industry 4.0"

The possibility of using machine learning methods to predict the results of high-voltage electric discharge treatment of titanium powder in a hydrocarbon liquid is studied. As a result of the work, distribution surfaces for the average particle diameter of Titanium powder. the amount of Titanium carbide formed during processing, and the number of spherical particles of titanium powder depending on the interelectrode gap and the number of pulses, when using spark discharge and with Titanium powder concentration in kerosene of 0.07 kg / dm3, pulse repetition frequency 0.3 Hz and the energy of single discharge of 1 kJ, were obtained.

Automatic classification of seismic sources finds vast application security domain. Due to restricted computational power in the edge devices a robust and less complex features need to be extracted and sent to appropriate classifier. Such features are histograms of oriented gradients. It represented the relative spectral distribution of derivative of amplitude against frequency instead of average spectral envelope.

The paper addresses the problem of acoustic signal classification and report the current state of the art in the field. More specifically the paper sheds light on the mathematical apparatus behind one specific method used for classification of sound sources – SVM (support vector machines). As well we propose use of histogram of oriented gradients (HOG) as features descriptor and explain how they work.

In this paper we will analyze what type of data we need in order to perform the research. Hence, we needed to gather a satisfying set of data that we will use for testing the initial model. In order to get to the right set of data, first it was crucial to identify the most common vineyard diseases in the Republic of North Macedonia. Furthermore, we needed to gather the data set from sensors from all available measuring stations in different vineyard locations. Finally, the aim is to develop an original model for vineyard disease prediction using the most appropriate machine learning algorithms. We present the latest results we have achieved as well.

In this paper, inertia wheel pendulum balance control is performed by using swing up and PID controller. Paper provides predictions on real time design balance system. Predictions were performed through data that were classified and tested by machine learning via MATLAB. Data obtained a result of the analyze of balance positions and swinging times of the wheel different diameters and weights in real-time. Through to this work will be able to predictable which wheel characteristics can be controlled and balanced

The main objective of this study is not to identify the best machine learning model, but instead to review the main datasets, publicly available, used to train and test security solutions that employ modern classification algorithms for anomaly detection. Hence, DARPA 1998 and KDD were studied as they were the first initiatives taken in this direction, while NSL-KDD, ISCXIDS2012 and CICIDS2017 are taken in consideration for future research because of their advantages. Personalized datasets will always bring a reasonable amount of uncertainty, especially since some feature vectors used for training remain unknown. Nevertheless, training on data specific to the protected infrastructure is more efficient, from the security point of view, than training on old attack signatures.

Nowadays, information systems security is a crucial aspect – vulnerable system endpoint can lead to severe data loss. Intrusion detection systems (IDS) are used to detect such unfortunate events. Implementation place defines the type of IDS: network-based (NIDS) for network traffic monitoring or host-based (HIDS), to detect malicious actions on the host level. IDS can be effective only if generated alerts are correctly evaluated and classified, what is typically done by a trained staff, but requires a lot of time and human resources. While a lot research is done with NIDS alerts evaluation, HIDS research is lacking behind. HIDS reported operating system calls could be used to define the importance of alarms and steer analysts to the most critical issues. In this article we demonstrate the applicability of our created Attack-Caused Windows System Calls Traces Dataset (AWSCTD), which is currently the most comprehensive dataset of system calls generated by almost all modern malware types, for training different classification methods on malware type recognition and later alert prioritization. The effectiveness of different classification methods is evaluated, and results are presented. Currently achieved results allow to decrease the load on analytical staff, dealing with malware classification and related alert prioritization by 92.4%, which makes this approach applicable for practical use.

The importance of materials in our life is well known to the world. Material science is one of the crucial research domain which deals with planning, development, synthesis and analysis of materials and its use in the real applications. The scientists and researcher are using highly advance technologies for the production and design of new materials in present era; however the process is time consuming. Machine Learning can be defined as a set of techniques that learn from the large amount of available data and make predictions for the new data. Recently, machine learning has gained a great attraction in material science research and has been used in several research studies. Machine learning techniques have the potential to learn from the large amount of materials data and make predictions about different properties for new materials of good quality. These predictions can help in designing new materials. In this paper, we present a comprehensive review on the machine learning applications in different material science domain. We believe that the paper would be useful for the researchers, academicians and students who are involved in the design of new materials.

The article suggests the consideration of data analysis ideology in the context of knowledge creation process and the technological patterns of social development. The problems of singularity (human misunderstanding of data processing results) associated with increase in data variety, volume and further intellectualization of the corresponding technologies for their processing are proposed to be solved by creating new formalization techniques that allow retransmission.