"MC-GAN Master Dataset"

"This dataset accompanies the research work titled Transformer-based Multi-Context GANs for Regime-Aware Synthetic Scenario Generation in Stocks. The dataset contains structured time-series data that integrates multiple contextual signals influencing stock market behavior, including sector-level financial indicators, macroeconomic variables like GDP, WPI, CPI and sentiment scores derived from Bombay Stock Exchange announcements. In addition, textual sentiment information extracted from BSE corporate announcements is included to provide contextual signals reflecting market sentiment. The dataset was created to help train and test Multi-Context Generative Adversarial Networks (MC-GAN), which create fake market situations based on economic conditions and other factors, thereby enabling researchers to simulate various market scenarios and analyze potential outcomes in stock market behaviour. "