Advanced Financial Time Series Anomaly Detection Framework

Design a comprehensive anomaly detection pipeline for financial time series data that can identify statistically significant market irregularities across multiple asset classes. The solution should incorporate z-score, IQR, and machine learning techniques like Isolation Forest. Specifically, develop a modular Python/Pandas architecture that can detect anomalies in trading volumes, price movements, and volatility indicators with configurable sensitivity thresholds. Include recommendations for handling false positives and integrate visualization techniques for interpretable results.