forecasting:meeting_minutes_october_25_2016

Differences

This shows you the differences between two versions of the page.

Link to this comparison view

forecasting:meeting_minutes_october_25_2016 [2016/10/26 02:48]
jobatake
forecasting:meeting_minutes_october_25_2016 [2021/09/19 21:59]
Line 1: Line 1:
-====== Forecasting Meeting Minutes for Week of October 18, 2016 ====== ​ 
-** Present Monday: Austin, Gordon, Brieanna, Jaimie ** 
  
-** Present Tuesday: Gordon, Brieanna, Jaimie ** 
- 
-===== Updates ===== 
-  * Sharif will meet with us on Wednesday 
- 
-===== Meeting with Dr. Kuh ===== 
-===== Expectations ===== 
-  * All of this is normalized (w/ Zenith Angle) 
-  * X' = [X / 1 1 1 1 1 1] 
-  * X'.T = [X.T | 1] 
-  * X'​*X'​.T = [__X*X.T__ | **X1**] 
-            * [1.T*X.T ​  | m     ] 
-  * (X'​*X'​.T)/​m 
-  *** Mean**: X1/m 
-  *__ Correlation(R)__ : X*X.T/m 
-  * Samples along diagonal of correlation matrix should be similar 
-    * Like toeplitz matrix 
-      * Diagonals same and symmetrical 
-  * Python programming help from Sharif tomorrow 
-  * MSE value 
-  * Root Mean Squared Working this week 
-    * Training to use same data for RMS 
-    * Go down with number of taps increasing 
-      * Take square root of the error 
-      * (90,000)^.5 = 300 
-  * Weight for taps 
-    * Positive values 
-    * Decrease in value 
-  * Train on 2011 and test on 2012 
-  * Train on 2012 and test on 2011 
  • forecasting/meeting_minutes_october_25_2016.txt
  • Last modified: 2021/09/19 21:59
  • (external edit)