This scholarly article investigates the use of intelligent-machine learning (I-ML) techniques to speculate on bubbles in stock markets, with a special emphasis on the Indian scenario. The study emphasizes the interrelationship between valuation, volatility, and volume of trade as critical indicators (3-Vs) to detect financial bubbles. The paper examines several notable financial bubbles from the twenty-first century, such as the Dot-Com, US Housing, and Cryptocurrency bubbles, to build a foundation for a bubble detection framework. In this research, we employ a multistep approach to analyze potential bubble phases in prominent Indian stocks. Technical indicators such as Relative Strength Index (RSI), Bollinger Bands (BB), Forward Price-to-Earnings (F: P/E) ratios are calculated to find out the disproportionate changes in 3-Vs. Subsequently, a new dataframe with a technical indicator is generated to train the long short-term memory (LSTM) model to predict potential bubble phases. The study carried out on Indian stocks shows that I-ML models, trained on technical indicators, reflect considerable effectiveness in forecasting bubble conditions and provide valuable insights for the early detection of market anomalies. The proposed I-ML model offers a practical solution for investors and regulators to mitigate risk of loss due to bubble formation in stock market. The paper concludes that the proposed approach using technical indicators along with real-time stock data to train deep models outperforms traditional time-series models and can open new avenues for economic forecasting in emerging markets.