An analyst working at production plant is asked to estimate the price of a chemical that is
Question:
An analyst working at production plant is asked to estimate the price of a chemical that is used in their production process. The chemical can be purchased from different suppliers. There are various attributes that quantify the quality of the particular chemical, and which may have an influence on its price. The analyst developed a linear regression model based these quality attributes in order to predict the price of the chemical. The linear regression modelling produced the following results:
Attribute | Coefficient | Std. Error | Std. Coefficient | Tolerance | p-Value |
Impurities | 5.383 | 2.201 | -0.549 | 0.162 | 0.019 |
Packaged weight | - 0.946 | 0.103 | -0.438 | 1.000 | 0.0 |
Density | 0.008 | 0.263 | 0.001 | 0.998 | 0.27 |
Reactive efficiency | -0.665 | 2.139 | -0.070 | 0.241 | 0.16 |
(Intercept) | 4116.427 | 60.806 | ? | ? | 0.0 |
(a) Write the regression equation for this model.
(b) Explain which attribute(s) do not play a statistically significant role in the above regression model at the 95% confidence level. Justify your answer.
(c) Explain which attribute(s) have a high likelihood of multi-collinearity in the above regression model. Justify your answer.
(d) The R-squared of the model was 0.729. Is this a good regression model? Justify your answer.
(e) Explain how you would remove the attribute(s) that are not statistically significant or have a high likelihood of multi-collinearity in the regression in a Rapid Miner process.
An Introduction to Management Science Quantitative Approach to Decision Making
ISBN: 978-1337406529
15th edition
Authors: David R. Anderson, Dennis J. Sweeney, Thomas A. Williams, Jeffrey D. Camm, James J. Cochran