What you'll learn

What is Hypothesis Testing? How to Frame Null & Alternative Hypothesis statements? What are the results of a hypothesis testing? And real-time examples.
Why do errors occur in hypothesis testing? What are Type I and Type II errors? How to choose Type I error ? How to balance between Type I & Type II errors?
How to use a point estimate to judge between Null & Alternative hypothesis? How to deploy Z or t statistic / table, calculate P-value, choose alpha and decide
What is Beta & Power of the hypothesis test? What factors influence Beta and Power of the test? How to estimate beta and power of the test?

Course content

5 sections • 5 lectures • 1h 32m total length

Introduction to Null & Alternative Hypothesis14:21
End of this lecture you will understand the answers to the following questions
Can a claim or a hypothesis around a population characteristic be judged based on a point estimate?
What is Null Hypothesis?
What is Alternative hypothesis?
How to frame the Null and the Alternative hypothesis under different scenarios?
Why Null hypothesis uses '=' symbol while alternative hypothesis uses one among the three {!=, > or <} symbols?
What is an Hypothesis test?
What are the possible results of an hypothesis test?
Students at the end of this lecture should be in a position to frame appropriate Null & Alternative hypotheses depending on the problem at hand

Type I Type II errors8:50
End of this lecture students will know the answers to each of these questions
Does hypothesis test result in errors? Is it normal?
What is Type I error?
What is type II error?
(understand the contextual meaning of each of the two errors via some real time examples)
How to assess the consequence of each of the above two errors based on the contest?
Are these errors independent of each other or one influences the other?
How to strike a balance between these two errors and limit each one of them based on the context?
At the end of this lecture the student will be in a position to assess the two errors and choose a suitable value for the type I error

Applying the concepts34:28
This is most important and a lengthy lecture in this playlist
This applies the concepts discussed in the previous two lectures and other essential statistical concepts together and explains step by step in detail and explain the decision making between the hypotheses.
Explains
how a point estimate (sample statistic) derived from the sample data is used to judge if the true population characteristic is close enough to the hypothesized mean or sufficiently in deviation from the hypothesized mean using sampling statistic distribution probabilities
how the alpha value is used to arrive at the z or the t critical value(s) depending the null and the alternative hypothesis under verification
how to derive the z or the t-statistic from the point estimate
how to calculate the P-value from the z or the t-statistic using the z or t-Statistic tables
How to compare P-value and alpha to arrive at the decision between null and the alternative hypothesis
All the above steps are explained for both forms of population characteristic the population mean and the population proportion for all possible forms of the alternative hypotheses involving (!=, >,<} inequalities

Illustration with examples15:27
This lecture is dedicated for illustration of the concepts applied in hypothesis testing discussed in lecture #3 using two real time examples
1. the first example involves testing an hypothesis for a population characteristic that is a population proportion of success
2. the second example involves testing an hypothesis for a population characteristic that is a population mean
Both these examples illustrate with data everything involving arriving at the point estimate, calculate z or t statistic using respective z or t-statistic tables, calculating the P-value, choosing a value for alpha, comparing P-value and alpha depending on the kind of the hypotheses statements under verification and arriving at a decision between rejecting the Null Hypothesis or not

Power of the Hypothesis test and Bets19:12
End of this lecture one will have the answers for
What is the definition of a good hypothesis test?
What is the power of the hypothesis test?
What is the power of the hypothesis test when H0 (Null hypothesis) is false?
What is the relation between Power of the test and probability of type II error (beta)?
What are the factors that influence beta (and thus the power of the test)?
How to visualize the above influences?
How to estimate beta and thus estimate the power of the test?

Requirements

It is good for the student to have a little understanding on sample distributions, point estimate, Z / t test statistic and tables

Description

Hypothesis testing is one of the most essential topics in any statistical study and machine learning algorithms

P-value, alpha and point estimates find their foot prints in most of the statistical and ML studies.

Hence it is important to understand hypothesis testing in as much detail as possible .

The course includes five lectures

Starting the first lecture with the concepts that help you understand Hypothesis testing and what Null & Alternative hypothesis are. How to frame Null & Alternative hypothesis. Explain with a few real time examples

Lecture 2 introduces you to the possible errors namely type I and Type II errors that occurs as a result of the hypothesis testing. Explains how to choose Type I error and balance between Type & II errors

The third lecture is the main lecture and an elaborate one. Helps you understand how to apply the concepts and carry out the hypothesis testing applying all the essential statistical concepts. Explains all the statistical steps and their sequence involved in carrying out the hypothesis test until the conclusion is arrived

The fourth lecture is dedicated for illustration of the hypothesis testing using real time examples

The fifth and the final lecture explains you all about the power of the hypothesis test. The roll of type II error and it probability beta on the power of the test

All the above steps Comprehensively cover all the details in great level of granularity that at the end of the course I am sure you will have a complete and a comprehensive understanding on Hypothesis testing and how to carry them out

#hypothesistesting #nullhypothesis #alternativehypothesis #pointestimate #pvalue #zstatistic #tstatistic #alpha #type1error #typeierror #beta #type2error #typeiierror #statistics #datascience

Who this course is for:

Useful for Data science aspirants and professionals and student communities across management, marketing, medical , economics, engineering and many other wide spectrum of disciplines

What you'll learn

Explore related topics

Course content

Introduction to Null and Alternative Hypotheses1 lecture • 14min

Type I & Type II errors1 lecture • 9min

Applying the concepts1 lecture • 34min

Illustration with examples1 lecture • 15min

Power of the Hypothesis Test and Beta1 lecture • 19min

Requirements

Description

Who this course is for: