RE:RE:RE:little predictive model
Sorry about the difficulty to read the text message as BB automatically collapsed my sentences and paragraphs. :( Barecode & Charlie_Chan: In my model, I used a time-series variable and a whole bunch of dummy variables. I checked the autocorrelation between SP (at close time) and the time-series variable, and I found a strong positive correlation. Note that, in security analysis, autocorrelation has an important application as mentioned by Investopedia: https://www.investopedia.com/terms/a/autocorrelation.asp The correlation partially explains why the adjusted R-spr is high. Skyhigh123: About the data set, to avoid confusion, I copied and repasted below the 2 data sets, one with actuals and the other one with prediction, so people can read and reuse the data. Actual data Date Actual_Val 10/09/2015 1.91 10/13/2015 1.88 ... 11/11/2015 2.63 11/12/2015 2.61 11/13/2015 2.70 11/16/2015 2.85 11/17/2015 2.86 11/18/2015 2.93 11/19/2015 2.96 11/20/2015 2.95 11/23/2015 2.87 11/24/2015 2.95 11/25/2015 3.18 11/26/2015 3.3 11/27/2015 3.37 11/30/2015 3.54 12/01/2015 3.14 12/02/2015 3.07 12/03/2015 2.95 12/04/2015 3.02 12/07/2015 3.22 12/08/2015 3.31 12/09/2015 3.12 12/10/2015 12/11/2015 Prediction Date Pred_Val 10/09/2015 1.67 10/13/2015 1.72 ... 11/11/2015 2.54 11/12/2015 2.61 11/13/2015 2.69 11/16/2015 2.78 11/17/2015 2.73 11/18/2015 2.78 11/19/2015 2.80 11/20/2015 2.90 11/23/2015 2.95 11/24/2015 2.93 11/25/2015 2.99 11/26/2015 3.04 11/27/2015 3.10 11/30/2015 3.18 12/01/2015 3.10 12/02/2015 3.19 12/03/2015 3.25 12/04/2015 3.27 12/07/2015 3.35 12/08/2015 3.34 12/09/2015 3.39 12/10/2015 3.45 12/11/2015 3.51 Hope this help. GLTA qh1234