Research on the Construction of Human-Computer Interaction System Based on a Machine Learning Algorithm

In this paper, we use machine learning algorithms to conduct in-depth research and analysis on the construction of human-computer interaction systems and propose a simple and effective method for extracting salient features based on contextual information. The method can retain the dynamic and static information of gestures intact, which results in a richer and more robust feature representation. Secondly, this paper proposes a dynamic planning algorithm based on feature matching, which uses the consistency and accuracy of feature matching to measure the similarity of two frames and then uses a dynamic planning algorithm to find the optimal matching distance between two gesture sequences. The algorithm ensures the continuity and accuracy of the gesture description and makes full use of the spatiotemporal location information of the features. The features and limitations of common motion target detection methods in motion gesture detection and common machine learning tracking methods in gesture tracking are first analyzed, and then, the kernel correlation filter method is improved by designing a confidence model and introducing a scale filter, and finally, comparison experiments are conducted on a self-built gesture dataset to verify the effectiveness of the improved method. During the training and validation of the model by the corpus, the complementary feature extraction methods are ablated and learned, and the corresponding results obtained are compared with the three baseline methods. But due to this feature, GMMs are not suitable when users want to model the time structure. It has been widely used in classification tasks. By using the kernel function, the support vector machine can transform the original input set into a high-dimensional feature space. After experiments, the speech emotion recognition method proposed in this paper outperforms the baseline methods, proving the effectiveness of complementary feature extraction and the superiority of the deep learning model. The speech is used as the input of the system, and the emotion recognition is performed on the input speech, and the corresponding emotion obtained is successfully applied to the human-computer dialogue system in combination with the online speech recognition method, which proves that the speech emotion recognition applied to the human-computer dialogue system has application research value.

Assessment of the risks of losing investments aimed at the development of Smart city systems

The article proposes a model for the computational core of the decision support system (DSS) in assessing the risks of investment loss during the dynamic planning (DP) of Smart City development. In contrast to the existing solutions, the proposed model provides specific recommendations when assessing the risks of loss. In case of an unsatisfactory risk forecast, it is possible to flexibly adjust the parameters of the investment process in order for the parties to achieve an acceptable financial result. The scientific novelty of the results is that for the first time it is proposed to apply a new class of bilinear multistep games. This class allowed us to adequately describe the process of assessing the risks of investment loss, using the example of dynamic planning for the placement of financial resources of players in Smart City projects. A distinctive feature of the considered approach is the use of tools based on the solution of a bilinear multistep game of both quality with several terminal surfaces, and a game of degree solved in the class of mixed strategies. Computational experiments were carried out in the Maple mathematical modeling package, and a DSS was developed in which a risk assessment model was implemented. The developed DSS allows to reduce the discrepancies between the data for predicting the risks of investment loss during the Smart City DP and the real return on investment.

Optimizing Inventory Replenishment for Seasonal Demand with Discrete Delivery Times

This study investigates replenishment planning in the case of discrete delivery time, where demand is seasonal. The study is motivated by a case study of a soft drinks company in Germany, where data concerning demand are obtained for a whole year. The investigation focused on one type of apple juice that experiences a peak in demand during the summer. The lot-sizing problem reduces the ordering and the total inventory holding costs using a mixed-integer programming (MIP) model. Both the lot size and cycle time are variable over the planning horizon. To obtain results faster, a dynamic programming (DP) model was developed, and run using R software. The model was run every week to update the plan according to the current inventory size. The DP model was run on a personal computer 35 times to represent dynamic planning. The CPU time was only a few seconds. Results showed that initial planning is difficult to follow, especially after week 30, and the service level was only 92%. Dynamic planning reached a higher service level of 100%. This study is the first to investigate discrete delivery times, opening the door for further investigations in the future in other industries.

SOLVING DYNAMIC PLANNING PROBLEMS ON THE BASIS OF GROUP DECISION METHODS

This article discusses problems based on group methods of decision-making based on a multimodal approach to solving the problems of dynamic planning of Smart City development. In particular, the main tool for decision-making in the process of dynamic planning is a system of models and methods, which are the basis for creating the structure of the dynamic plan, optimizing the plan and finding the desired solution. Collective decision-making methods are used in the collective development of planning decisions, including in the context of solving the problems of dynamic planning of Smart City development and group decision-making, which requires the agreed decisions of several specialists.

The Accuracy of Patient-Specific Instrumentation with Laser Guidance in a Dynamic Total Hip Arthroplasty: A Radiological Evaluation

The functional positioning of components in a total hip arthroplasty (THA) and its relationship with individual lumbopelvic kinematics and a patient’s anatomy are being extensively studied. Patient-specific kinematic planning could be a game-changer; however, it should be accurately delivered intraoperatively. The main purpose of this study was to verify the reliability and accuracy of a patient-specific instrumentation (PSI) and laser-guided technique to replicate preoperative dynamic planning. Thirty-six patients were prospectively enrolled and received dynamic hip preoperative planning based on three functional lateral spinopelvic X-rays and a low dose CT scan. Three-dimensional (3D) printed PSI guides and laser-guided instrumentation were used intraoperatively. The orientation of the components, osteotomy level and change in hip length and offset were measured on postoperative CT scans and compared with the planned preoperative values. The length of surgery was compared with that of a matched group of thirty-six patients who underwent a conventional THA. The mean absolute deviation from the planned inclination and anteversion was 3.9° and 4.4°, respectively. In 92% of cases, both the inclination and anteversion were within +/− 10° of the planned values. Regarding the osteotomy level, offset change and limb length change, the mean deviation was, respectively, 1.6 mm, 2.6 mm and 2 mm. No statistically significant difference was detected when comparing the planned values with the achieved values. The mean surgical time was 71.4 min in the PSI group and 60.4 min in the conventional THA group (p < 0.05). Patient-specific and laser-guided instrumentation is safe and accurately reproduces dynamic planning in terms of the orientation of the components, osteotomy level, leg length and offset. Moreover, the increase in surgical time is negligible.