- Difficulty:

- Terrain:

Size: (small)

**Cache Issues:**

- This cache is temporarily unavailable. Read the logs below to read the status for this cache.

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.

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(e^{2}[n]) is given by the following equation:

**c **= **R _{xx}^{-1} p**,

where **R _{xx}** is the autocorrelation matrix of the input process x[n], and

-- Let AB be the rank of the matrix **R _{xx}** (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:

Logbook

Dr. Dolittle

Marge

Bronze/Silver/Golden Breastplate of the Adapter (Bonus Trading Cards for FTF/STF/TTF)

**
Additional Hints**
(Decrypt)

Hagre-Ubym

Haqre-Jbbq

Haqre-Jbbq

Decryption Key

A|B|C|D|E|F|G|H|I|J|K|L|M

-------------------------

N|O|P|Q|R|S|T|U|V|W|X|Y|Z

Find...

59 3 2 2 1 1

View Logbook | View the Image Gallery

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

Current Time:

Last Updated:

Rendered From:Unknown

Coordinates are in the WGS84 datum

- Get to Know Us
- Careers
- Partner With Us

- Shop Geocaching
- USA/Canada Shop
- International Retailers

- Contact Us
- Help Center
- Media Inquiries

Choose Language

© 2000-2017 Groundspeak, Inc. All Rights Reserved. Groundspeak Terms of Use | Privacy Policy