How do I do data analysis?

Data Science, AI and ML
#1
Although there may be certain specific ways of catering to specific business problem; but the generic view of data analysis invariably comprises five key steps – firstly, it starts with defining the business problem as all statistical or data solutions should always correspond to business problems; second, is to set-up measurement criteria i.e. to convert the business problem into a technical or statistical problem; then comes the data collection phase which could be via both primary or secondary mechanisms.
Once the data is collected, cleaned & harmonized, the fourth step is to analyse the data using various tools & technologies; and finally, interpret the inferences from the analysis and communicate the result to the business users.
#2

Data analysis depends on the data you are looking into, where the overall process remains same for any type of data. The following steps gives clear idea on how to process and interpret data,

  1. Need to decide the test statistic (the why part)
  2. Now decide the best source for required data
  3. Get the data from the source for required time interval and do sanity check of the data
  4. Data cleaning - To use the data for perfect analysis clean the data to make things easier
  5. Analyze the data for by several interpretation and Visualization techniques
  6. Interpret the results and draw conclusions from historic data