The purpose of the theoretical considerations and research conducted was to indicate the instruments with which the quality of a dataset can be verified for the segmentation of observations occurring in the dataset. The paper proposes a novel way to deal with mixed datasets containing categorical and continuous attributes in a customer segmentation task. The categorical variables were embedded using an innovative unsupervised model based on an autoencoder. The customers were then divided into groups using different clustering algorithms, based on similarity matrices. In addition to the classic k-means method and the more modern DBSCAN, three graph algorithms were used: the Louvain algorithm, the greedy algorithm and the label propagation algorithm. The research was conducted on two datasets: one containing on retail customers and the other containing wholesale customers. The Calinski–Harabasz index, Davies–Bouldins index, NMI index, Fowlkes–Mallows index and silhouette score were used to assess the quality of the clustering. It was noted that the modularity parameter for graph methods was a good indicator of whether a given set could be meaningfully divided into groups.

Purpose:The aim of the article is to develop an algorithm for forecasting sales in the supply chain based on the LSTM network using historical sales data of a furniture industry company. Design/Methodology/Approach: Machine learning was used to analyze the data. The method of predicting the behavior of sales value in a specific time horizon in terms of a time series was presented. The LSTM network was used to predict sales. The network used is a special case of recursive neural networks with an important difference in the repeating module. Due to the fact that the activities are carried out on time series, the data was analyzed in terms of the stationarity of such series or trends and seasonal effects. The data used in the analysis includes the daily sales values of a group of certain furniture collections over a specified time horizon. The stationarity of the time series can have a significant impact on its properties and behavior prediction, where failure to bring the time series to the correct form of stationarity can lead to false results. Findings: The result of the research was the analysis of sales forecasting in the supply chain based on machine learning. As a result of the data transformations, the algorithm was able to recognize and learn long-term relationships. Practical Implications: The presented method of predicting the behavior of sales value in a specific time horizon allows for a look at the forecasting of demand in terms of the supply chain. The sales data of a company from the furniture industry were used for the analysis. Originality/Value: A novelty is the use of the LSTM network trained on real transaction data of a furniture company that has based its business on the supply chain and cooperates with its suppliers and recipients in Central and Eastern Europe.

No abstract

This paper presents a novel, autonomous learning system working in real-time for face recognition. Multiple convolutional neural networks for face recognition tasks are available; however, these networks need training data and a relatively long training process as the training speed depends on hardware characteristics. Pretrained convolutional neural networks could be useful for encoding face images (after classifier layers are removed). This system uses a pretrained ResNet50 model to encode face images from a camera and the Multinomial Naïve Bayes for autonomous training in the real-time classification of persons. Faces of several persons visible in a camera are tracked using special cognitive tracking agents who deal with machine learning models. After a face in a new position of the frame appears (in a place where there was no face in the previous frames), the system checks if it is novel or not using a novelty detection algorithm based on an SVM classifier; if it is unknown, the system automatically starts training. As a result of the conducted experiments, one can conclude that good conditions provide assurance that the system can learn the faces of a new person who appears in the frame correctly. Based on our research, we can conclude that the critical element of this system working is the novelty detection algorithm. If false novelty detection works, the system can assign two or more different identities or classify a new person into one of the existing groups.

No abstract