studying the peculiarities of loneliness among refugees and internally displaced persons due to the Russian-Ukrainian warand forecasting migration processes

Responsibilities:

• Analysis of loneliness indicators as influential factors; determining correlation coefficients between loneliness and studied indicators.

• Analysis of individual categories of loneliness factors.

• Development of a forecasting model.

• Development of a dashboard for visualizing key research metrics.

Results:

• Conducted statistical data processing using descriptive statistics, Student's t-test,Mann-Whitney U-test, correlation and factor analysis.

• Analyzed data for 2022 to study the structure of loneliness among women forcedmigrants, identifying significant characteristics affecting the likelihood of their return to Ukraine.

• Confirmed the optimal number of factors using Kaiser criterion and scree plot with a factor loading threshold of 0.7

• Developed a model for predicting the likelihood of return based on previously identified factors.

• Created an interactive dashboard to present analysis results and use the forecasting model to interested parties using Dash and Plotly.

Responsibilities:

• Data collection and preparation; analysis of biometric and user data quality.

• Development and tuning of machine learning models.

• Performance analysis of algorithms.

• Creation of visualizations and reports.

Results:

• Developed and implemented computer vision models for a biometric anti-fraud system.

• Conducted exploratory data analysis and prepared image datasets for model training.

• Used PyTorch to build and train neural networks.

• Achieved classification model accuracy exceeding 90%.

• Analyzed classifier quality using Precision, Recall, and F1-Score metrics.

• Conducted A/B testing to confirm the best model using Student's t-test for in dependent samples and Wilcoxon test for dependent samples.

• Created visualizations of computer vision model results using OpenCV, Matplotlib, and Seaborn.

Responsibilities:

• Form quarterly analytical reports: profitability, demand and supply.

• Provide relevant database extracts for the ERP task force, which reduces enterprise resource planning and conducts competitive analysis.

• Research the primary and consolidated documentation of the enterprise for monitoring progress & preparation of visualization of the main business metricsin dynamics.

• Analyze the effectiveness of sales managers.

Results:

• Created an analytical report on construction product demand for 2016—2017 using SQL and Excel to manage inventory.

• Based on a comprehensive SWOT-analysis of the management team, identified leading sales channels and regions of major competitors, pinpointing areas where sales could be improved with targeted marketing efforts.

• Created presentations and infographics of absolute and relative business metrics dynamically using Excel and PowerPoint.

• Implemented checklists for routine tasks of individual company departments. Based on hypothesis testing results, developed a sales incentive program.