Generalized Multi-Timeframe Probability Estimation

[Tim55667757]

After implementation of #145

🎯 Goal:

Develop a generalized function to estimate the probability of reaching a target price using more than two timeframes simultaneously. This is an extension of the existing EstimateTargetReachability() method, which currently supports only two time series.

---

📌 Problem Statement

The current implementation of EstimateTargetReachability() supports dual-timeframe estimation using log returns, volatility, and a Bayesian + weighted aggregation mechanism. However, the aggregation logic is not associative, meaning we cannot apply it recursively or incrementally across multiple timeframes.

To support 3 or more timeframes (e.g., 5m, 1h, 4h, 1d), we must design a new function that:

• Handles N independent price series, each with its own forecast horizon.
• Computes individual probabilities p₁, p₂, ..., pₙ using the same log-return and volatility framework.
• Aggregates these using a proper global strategy, avoiding pairwise recursion.

---

🧠 Key Design Considerations

• The aggregation must account for:
  • Individual volatility levels (σ₁, σ₂, ..., σₙ)
  • Horizon in candles for each timeframe
  • Directional confidence alignment (e.g., high volatility vs. stable TFs)
• Output format should be consistent with current system:

  def EstimateTargetReachabilityMulti(...) -> Tuple[float, str]