Implementation of PC A using Matlab

Ditutup Disiarkan Nov 27, 2015 Dibayar semasa penghantaran
Ditutup Dibayar semasa penghantaran

The goal of this project is to determine whether PC A as implemented by Matlab can

successfully un-mix a known combination of signals to obtain principle components.

1. Construct a set of three sinusoidal signals of different frequencies that are

not harmonics of one another – for each data set, add a small amount of

white noise. (SNR of less than 5%)

2. Now create three signals that are different linear combinations of the three

sinusoidal frequencies.

3. Using the Matlab PCA command, determine the coefficient matrix for the PCA

transformation.

4. Using the coefficient matrix, determine the transformed data set along

principal component axes.

5. Plot the noisy sinusoids, from part 1, the data sets from part 2, and the result

from part 4.

6. In your project write-up address whether or not PCA successfully separated

the signals. Why or why not?

Example

To clarify:

Construct three sinusoidal signals of different frequencies. Say the signals are called x1, x2, and x3. Choose your frequencies to be within an order of magnitude, but not harmonics. So, say 5 hertz, 12 hertz, 17 hertz.

Add some random noise to each signal.

Now, create three new signals that are linear combinations of x1, x2, and x3. Let's call them f1, f2, and f3

So, something like f1 = 3x1 + 7x2 + 4x3 and so on.

The "f" signals are the ones you will run PCA on. The objective is to see whether PCA will recover your original signals.

As part of the project presentation you should:

plot your original x signals

plot your f signals

plot your signals that PCA gives you

Your output from PCA will be a matrix of coefficients. When the PCA matrix multiplies the f signals, it is effectively "unmixing" the signal and giving you the principal components. Your goal is to determine how effective PCA is on the signals you create.

Matlab and Mathematica

ID Projek: #8986610

Tentang projek

7 cadangan Projek jarak jauh Aktif Jan 3, 2016

7 pekerja bebas membida secara purata $372 untuk pekerjaan ini

ognjen011

Hello! This can be done quikly and comprehensively, if you choose me. Ping me for further discussion. All the best! Ognjen

$250 USD dalam 3 hari
(97 Ulasan)
6.3
elephentplayer

Hello, dear friend I am expert in matlab, mhthematics, physics,... I have a lot of experiences in PCA. I can help you well. Please contact me. Best regards

$250 USD dalam 10 hari
(61 Ulasan)
5.2
kjgroup197318

---------------------------------Nice job.I can do it. I'm math expert. No problem.---------------------------

$750 USD dalam 7 hari
(21 Ulasan)
4.5
x3x7

Quite detailed description, I guess most of the time I allocated will be spent on synchronisation with a chat or e-mail about the actual results than work itself :) Kind regards, x3x7 p.s. if you have combined l Lagi

$250 USD dalam 3 hari
(1 Ulasan)
4.0
bittersweet00

Hi, I'm electrical and electronics engineer. I work in signal processing task as R&D Engineer. I use Matlab for 4 years. I can start immediately. Best wishes, Murat

$300 USD dalam 4 hari
(2 Ulasan)
2.2
anacris1

Ph.D. physicist with extensive expertise in scientific computation, mathematical modeling, Monte Carlo and stochastic simulations, quantitative analysis, software design and algorithm development. Past areas of researc Lagi

$526 USD dalam 10 hari
(0 Ulasan)
0.0
EXPERTRADING2

DEAR HIRING MANAGER I am an industrial engineer. I am very good at mechanical, electrical, control and automation, fluid mechanical,...I am expert at system modeling,data base creating,data quantitative analysis,sta Lagi

$277 USD dalam 3 hari
(0 Ulasan)
0.0