Skip to Content

Mystery Cache


A cache by sigurd_fjoelskaldr Send Message to Owner Message this owner
Hidden : 11/16/2013
In Steiermark, Austria
4.5 out of 5
1.5 out of 5

Size: Size: small (small)

Join now to view geocache location details. It's free!


How Geocaching Works

Related Web Page

Cache Issues:

  • This cache is temporarily unavailable. Read the logs below to read the status for this cache.
Please note Use of services is subject to the terms and conditions in our disclaimer.

Geocache Description:

This cache draws you into the exciting world of adaptive linear filtering - a tough task, but manageable with a little linear algebra and probability theory. I hope you enjoy it!

Information about the cache (and ways to solve it) can be found in the link below or in the book "Adaptive Filter Theory" by Simon Haykin.

Block Diagram

The first goal is to solve a mathematical problem: Consider the figure above, where c is an adaptive filter with three coefficients, i.e., c=[C,D,E]. h is a linear, time-invariant filter with two coefficients, h=[5,5]. x[n] is a unit-variance white noise process with zero mean. v[n] is jointly stationary with x[n], with cross-correlation

E(x[n-k]v[n]) = 2 if k=2 and 0 else.

The filter coefficient vector minimizing the mean-squared error E(e2[n]) is given by the following equation:

Rxx-1 p,

where Rxx is the autocorrelation matrix of the input process x[n], and p is the vector of cross-correlations between x[n] and d[n].


-- Let AB be the rank of the matrix Rxx (e.g., AB=07 or AB=15).

-- Compute the coefficient vector c=[C,D,E].

-- This theory behind this problem is intricately linked to a famous mathematician, who also investigated the mathematical aspects of Brownian motion and is considered the founder of cybernetics. He was born on DD-MM-YYYY; FG=DD+1.

-- It is also possible to get the solution without knowing the process statistics, by letting an algorithm run. This algorithm just takes measurements of the input process x[n], the desired output d[n], and the error e[n], and changes the coefficient vector c such that the squared error is minimized. The algorithm was developed by a professor and his PhD student. H is the second letter of the professor’s first name (e.g.: Sigurd, second letter = i, H=9). Hint: The professor also contributed greatly to the theory of quantization.

-- To which of the following applications of adaptive filters does the example above belong?

I=1: inverse modeling
I=2: prediction
I=3: system identification
I=4: interference cancellation

-- The matrix equation solving the problem above is also linked to an Austrian mathematician. He died in 1983. J = digit sum of his age - 3.

Please take care about muggles hiding behind windows! I recommend grabbing the cache during the night or on weekends. Very important: The cache is NOT connected in any way to the measurement device (Feuchtemessung), so please do not touch the latter!
Original content:
Dr. Dolittle
Bronze/Silver/Golden Breastplate of the Adapter (Bonus Trading Cards for FTF/STF/TTF)

Additional Hints (Decrypt)


Decryption Key


(letter above equals below, and vice versa)



68 Logged Visits

Found it 59     Didn't find it 3     Write note 2     Temporarily Disable Listing 2     Enable Listing 1     Publish Listing 1     

View Logbook | View the Image Gallery

**Warning! Spoilers may be included in the descriptions or links.

Current Time:
Last Updated: on 9/17/2017 11:09:09 PM Pacific Daylight Time (6:09 AM GMT)
Rendered From:Unknown
Coordinates are in the WGS84 datum

Return to the Top of the Page