High Points in Time Series Data

Requirements Clarification & Assessment

1. Objective Understanding:

   • The task involves detecting high points or peaks in a time series dataset characterized by fluctuations.
   • Clarify what constitutes a "high point." Is it a local maximum, a point above a certain threshold, or a statistical anomaly?
   • Determine if both positive and negative peaks need detection.

2. Data Characteristics:

   • Understand the nature of the time series: Is it stationary or non-stationary?
   • Are there any known trends or seasonality?
   • What is the frequency of data collection (e.g., daily, hourly)?

3. Performance Criteria:

   • Define what "good" detection means: precision, recall, or a balance of both?
   • Is real-time detection required, or is batch processing sufficient?

4. Technical Constraints:

   • Are there computational constraints, such as memory or processing power?
   • Do you have access to historical data for model training or parameter tuning?

5. Output Requirements:

   • What format should the output be in? (e.g., indices of peaks, annotated time series)
   • Should the solution provide confidence scores or just binary peak detection?