Rajireddy Gudipati

This study analyzes the evolution of employee skill demand in the United States using data science and natural language processing techniques. Data was collected from multiple sources, including the US Federal Government Job Postings API, Federal Payroll Data, National Labor Statistics, and New York City open data. Analytical methods such as Natural Language Processing (NLP), Latent Dirichlet Allocation (LDA) topic modeling, Dynamic Topic Modeling, trend analysis, geospatial analysis, and embedding models were applied to identify patterns in skill demand. The results highlight major skill groups related to aviation and safety, engineering and human resources, and medical fields. The analysis also shows connections between skill demand, unemployment trends, and wage changes, particularly during the COVID-19 period. Overall, the study demonstrates how skill demand varies across time, regions, and levels of employment.

There have been raising global economic shocks that have necessitated the realization of recovery dynamics by policymakers, and economic planners. This project uses a time series model to predict how the forecast the speed and quality of economic recovery in the aftermath of enormous shocks like the 2008 financial crisis and the coronavirus pandemic. On the basis of several macroeconomic indicators such as GDP per capita, inflation, employment, population, and dependency ratios, the study compares several forecasting models such as XGBoost, ARIMA, Holt Exponential Smoothing, and Long Short-Term Memory (LSTM). The performance of the models is compared against MAE, RMSE, MAPE, MASE and Directional Accuracy. Also, soft vote ensemble model is adopted in order to enhance predictive robustness. Findings show that a combination of statistical, machine learning, and deep learning models is a more holistic framework of predicting a range of economies in terms of economic recovery.