Data Analytics for Forecasting Labor Market Dynamics in Bangladesh: A Category-wise Analysis

This study investigates the application of clustering and time series forecasting techniques to analyze sectoral trends in Bangladesh’s job market, specifically focusing on the IT/Telecommunications and Finance/Accounting sectors. We apply the K-means clustering algorithm to group job market data based on key variables such as industry growth rates, employment trends, and regional factors. Additionally, ARIMA is used for time series forecasting to predict sectoral growth trends from 2024 to 2030. The methodology begins with data preprocessing, followed by clustering to identify patterns, and then ARIMA for forecasting future trends. A comparison of the performance of K-means clustering under different numbers of clusters reveals its impact on the clarity and accuracy of the sectoral groupings. We further analyze the effect of varying the number of clusters on forecasting accuracy, demonstrating the trade-offs between cluster granularity and forecast reliability. Our findings show that the K-means algorithm’s performance improves with an optimal number of clusters, while ARIMA provides valuable insights into future sectoral growth. The practical implications of these results emphasize the need for accurate sectoral classification and forecasting in decision-making, taking into account market dynamics such as technological advancements, policy changes, and regional variations. Future work should explore the integration of qualitative factors and assess the long-term sustainability of the observed growth trends.