0:02 analysis is the process of gathering
0:05 cleaning analyzing and mining data
0:08 interpreting results and reporting the findings
0:09 findings
0:12 with data analysis we find patterns
0:14 within data and correlations between
0:17 different data points and it is through
0:19 these patterns and correlations that
0:21 insights are generated and conclusions
0:23 are drawn
0:25 data analysis helps businesses
0:27 understand their past performance and
0:30 informs their decision making for future actions
0:31 actions
0:33 using data analysis businesses can
0:36 validate a course of action before
0:38 committing to it saving valuable time
0:41 and resources and also ensuring greater success
0:42 success
0:45 we'll explore four primary types of data
0:48 analysis each with a different goal and
0:50 place in the data analysis process
0:53 descriptive analytics helps answer
0:56 questions about what happened over a
0:59 given period of time by summarizing past
1:00 data and presenting the findings to stakeholders
1:02 stakeholders
1:04 it helps provide essential insights into
1:06 past events
1:09 for example tracking past performance
1:10 based on the organization's key
1:13 performance indicators or cash flow analysis
1:14 analysis
1:17 diagnostic analytics helps answer the
1:19 question why did it happen
1:22 it takes the insights from descriptive
1:24 analytics to dig deeper to find the
1:27 cause of the outcome for example a
1:29 sudden change in traffic to a website
1:32 without an obvious cause or an increase
1:34 in sales in a region where there has
1:36 been no change in marketing predictive
1:39 analytics helps answer the question what
1:42 will happen next historical data and
1:44 trends are used to predict future
1:46 outcomes some of the areas in which
1:49 businesses apply predictive analysis are
1:52 risk assessment and sales forecasts it's
1:54 important to note that the purpose of
1:57 predictive analytics is not to say what
2:00 will happen in future its objective is
2:02 to forecast what might happen in the
2:05 future all predictions are probabilistic
2:07 in nature
2:09 prescriptive analytics helps answer the
2:12 question what should be done about it
2:15 by analyzing past decisions and events
2:17 the likelihood of different outcomes is
2:19 estimated on the basis of which a course
2:21 of action is decided
2:23 self-driving cars are a good example of
2:26 prescriptive analytics they analyze the
2:28 environment to make decisions regarding
2:31 speed changing lanes which route to take
2:34 etc or airlines automatically adjusting
2:37 ticket prices based on customer demand
2:40 gas prices the weather or traffic on
2:42 connecting routes now let's look at some
2:45 of the key steps in any data analysis process
2:46 process
2:49 understanding the problem and desired
2:51 result data analysis begins with
2:53 understanding the problem that needs to
2:56 be solved and the desired outcome that
2:59 needs to be achieved where you are and
3:01 where you want to be needs to be clearly
3:04 defined before the analysis process can begin
3:06 begin
3:08 setting a clear metric
3:09 this stage of the process includes
3:12 deciding what will be measured for
3:15 example number of product x sold in a
3:18 region and how it will be measured for
3:20 example in a quarter or during a
3:23 festival season gathering data once you
3:25 know what you're going to measure and
3:27 how you're going to measure it you
3:30 identify the data you require the data
3:32 sources you need to pull this data from
3:35 and the best tools for the job
3:38 cleaning data having gathered the data
3:41 the next step is to fix quality issues
3:43 in the data that could affect the
3:45 accuracy of the analysis
3:47 this is a critical step because the
3:49 accuracy of the analysis can only be
3:52 ensured if the data is clean
3:54 you will clean the data for missing or
3:56 incomplete values and outliers for
3:59 example a customer demographics data in
4:02 which the age field has a value of 150
4:05 is an outlier you will also standardize
4:08 the data coming in from multiple sources
4:11 analyzing and mining data
4:13 once the data is clean you will extract
4:15 and analyze the data from different
4:18 perspectives you may need to manipulate
4:20 your data in several different ways to
4:22 understand the trends identify
4:25 correlations and find patterns and variations
4:27 variations
4:29 interpreting results
4:31 after analyzing your data and possibly
4:33 conducting further research which can be
4:36 an iterative loop it's time to interpret
4:38 your results
4:40 as you interpret your results you need
4:42 to evaluate if your analysis is
4:44 defendable against objections and if
4:46 there are any limitations or
4:49 circumstances under which your analysis
4:51 may not hold true
4:53 presenting your findings
4:56 ultimately the goal of any analysis is
4:59 to impact decision making the ability to
5:01 communicate and present your findings in
5:04 clear and impactful ways is as important
5:07 a part of the data analysis process as
5:10 is the analysis itself reports
5:13 dashboards charts graphs maps case
5:15 studies are just some of the ways in
Chrome Extension