Question

How are exploratory data analysis (EDA) and hypothesis testing different? Explain why EDA could be preferred in data mining, and justify your explanation with a specific example.

Answer 1

Answer)

Hypothesis analysis would have the specifics with respect to the analysis one would perform and wish to perform. It is where the null hypothesis is being stated and it is where the data would get collected in a repeatable way and would also result in sampling the design while involving the random, or regular distribution that the study would plot. In case there would be any subjective involvement of the locating or the orienting study plots then it would result in technically not a valid response. It is where the analysis and variation would have the data transformation as well as the use of the different ordination option that would be planned else it is the user which would run the risk of data diving or data mining that would significantly result as there are many such options that have been tried. There are even stepwise techniques that would automate the forms of the data and would also lead to incorrect statistical inference. It is the reward for rigorous adhering for the stringent criteria which could help in statistical inferences that are turned as valid.

The exploratory analysis would lack the statistical rigor that would make the most for the vegetation research and the reason for the exploratory analysis which can help in finding the patterns that would inherently be subjective to the enterprise where the analysis would help in incorporating the skills, intuition of the investigator as well as wisdom when an experiment is considered. When one would find the other investigator with the identical wisdom or other fields then it is the analysis that is not being strictly repeatable and hence it is not even falsifiable. Also, it is even possible for performing the exploratory analysis and that too on a sample plot that is located as per the objective sampling design, a careful placement that is not necessary and it aides the investigator for subjectively placing the study plots in the considered location where one would consider it to be important as well as interesting

EDA could very well be used in the data mining as it is considered as an important step that would allow one to achieve the defined insights as well as statistical measures which would be essential for any business continuity, stakeholder as well as the data scientists for performing one to define as well as refine the related important feature for the valuable selection that can be used in one model

Example: Box and Whisker plots

If you found the answer useful, hit like
Hope this answer helps.
Thanks