Exponential Moving Average filter

แชร์
ฝัง
  • เผยแพร่เมื่อ 26 พ.ย. 2024

ความคิดเห็น • 7

  • @mathewkristensen3589
    @mathewkristensen3589 6 หลายเดือนก่อน

    Terrific explanation. Would love a series where you go over other filters like you mentioned with the iceberg.

  • @jeremyhuang4988
    @jeremyhuang4988 8 หลายเดือนก่อน

    Great video! Very well made

  • @andrewh2341
    @andrewh2341 8 หลายเดือนก่อน

    It would be interesting to review the different windowing options we can apply to discrete time sampled data :)

    • @gergelybencsik8626
      @gergelybencsik8626  8 หลายเดือนก่อน +1

      Do you mean the Hamming window and the like?

  • @zeKite
    @zeKite 5 หลายเดือนก่อน

    Was going to suggest Kalman but then i saw it at the bottom of the ocean. I think Filters that are easy to implement are the best in most scenarios

    • @gergelybencsik8626
      @gergelybencsik8626  5 หลายเดือนก่อน

      Indeed, and I find the 1st order lowpass to be the simplest, and it gets the most use. Kalman Filter is powerful, but overrated. I'd love to get there eventually, but sadly I don't have time for videos these days

  • @wolpumba4099
    @wolpumba4099 8 หลายเดือนก่อน

    *Abstract:*
    The video provides a comprehensive overview of the Exponential Moving Average (EMA) filter, a fundamental tool in signal processing. The presentation covers the filter's basic properties, implementation, and applications.
    *Key Points:*
    * *Definition and History:* The EMA filter is a first-order, low-pass filter with origins in 19th-century mathematics. It calculates the output (y) by recursively combining the current input (u) and previous output (y) with complementary coefficients (alpha and 1-alpha).
    * *Tuning:* The time constant (tau), which is the inverse of alpha, is a more intuitive parameter for tuning the filter's behavior.
    * *Implementation:* The EMA filter can be easily implemented in code due to its recursive nature.
    * *Properties:*
    * *Delay:* Causes a delay between 1 and 3 time constants.
    * *Cutoff Frequency:* The cutoff frequency is the inverse of the time constant (tau).
    * *Noise Reduction:* Reduces white noise with a factor related to the time constant.
    * *Comparison to Simple Moving Average (SMA):* While both filters offer similar noise reduction and delay properties, the EMA has advantages in memory usage, ease of implementation, and adaptability.
    i used gemini