The article reviews if the use of Big Data is a big mistake or not. Big data can be really helpful for the latest trends in the field of business and science and technology. Big data is really helpful for the companies for their expansion and growth.
As these companies have extremely large data set. In comparison to the size involved, the price is really less. The updating of the data is really simple. The results produce using data analysis are very accurate. There is a provision to find a co relation between the data. The companies like Amazon, Google and Facebook try to develop new method for understanding the processes by the help of data.
The potential of the big data needs to be used to the utmost limit to get the desired results. However, there are some problems which can occur in the analysis of the big data which must be avoided.
Ans 2.
The equation shows the linear dependence of the daily stock market return Rt in terms of various factors in a linear fashion. The values of various parameters obtained when the data is fit in a given equation have been mentioned in the table given. It shows a positive dependence on the variable like C, S, T, P and the negative dependence on the variables like MON, TUE, THU, E. the value of standard error, t statistics and probability have been mentioned for each of the variables. The value of R quad is denoting the closeness of the data to the regression line fitted. The value of the f statistics shows the ratio of the two quantities which are expected to be equal under the null hypothesis. The standard error of regression has a high value of around 0.9. the f statistics hold the value of 11.4.
Ans 3.
In the above task the sample size is large for a long period of time. The p value is very less irrespective of the economic significance of the effect size that can bias the outcome of the statistical inference. Th structural changes can also lead to distortion in the results. The regression model is estimated for a different sub sample for twenty years. First, Bayes factor for H0:
B5=0 for H1:
B5/= 0
The bays factor is given by the ration of likelihood of a given hypothesis to the likelihood of the other. In this case it is -22507.30.
RESET
The RESET test is used for the linear regression model in the given case the linearity is very less. Given by 0.0054.
Test for non-normality, heteroskedasticity, autocorrelation.
The Durbin Watson test can be used for the test of autocorrelation. Th statistic value is 2 in this case.
