## Lecture: Causal Networks with Independence
## Assigned Reading**Session Overview****Network Concept**- Read Chapter 20 in Statistical Analysis of Electronic Health Records, pages 487 to 497
- What is a cause and how is it different from correlation Slides►
- Displaying a causal network Slides► Video► YouTube►
- Causal chain, common effect, and common cause in 3 variables Slides► YouTube► Video►
- Parents, children and cycles in networks Slides►
- Markov Blanket Slides► PubMed►
- Markov blanket visualization Web►
- Fidelity between network displays and formulas Slides► Video► YouTube►
- Propagation through a network (using SQL) Slides► YouTube► Video►
- Learning Markov blankets using Grow-Shrink algorithm Slides► YouTube► Video►
- Impact of timing of variables on learning network structure Read► R code► Slides► Solu's Teach One YouTube►
- Optimizing stratification Read►
## AssignmentInclude in the first page a summary page. In the summary page write statements comparing your work to answers given or videos. For example, "I got the same answers as the Teach One video for question 1." For these assignment you can use any statistical package, including R, SAS, and SPSS, Python. R packages and BNLearn are also used often. OpenBUGS and Gibbs Sampler, Stan, OpenMarkov, and Direct Graphical Model are also open source software. Netica is free for networks less than 15 nodes.
Resources for Question 1: - Gelila Aboye's Teach One Slides► YouTube►
- Wang's Teach One, YouTube►
- Sully's Teach One, Slides►
- Sully's Python Code, see notes in Slides►
Additional resources for Question 2: - Data Download►
- Arshi Moktader's Teach One Slides► YouTube►
- Bushra's Teach One Read►
- Anto's Teach One YouTube►
- Slides►
- SQL►
Resources for Question 3: - Netica software Download►
- Nikhitha Telagamalla's Teach One YouTube► Slides (with Embedded Movie)►
- Shruti's Teach One using SQL YouTube►
- Usman's Teach One using SQL YouTube►
- Sully's Teach One, Python code is in notes: Slides►
- If X1 and X2 are significant predictors of Y, X3 and X4 are not, and no interactions are significant; then what is the Markov Blanket for Y? How is the concept of Markov Blanket related to multi-colinearity?
- Suppose X2 occurs after Y and X1 occurs prior to Y, what is a Markov Blanket that separates variables that are irrelevant and could possibly be causes of Y. Keep in mind that a cause is something that occurs prior to effect, has a significant association with the effect, has a mechanism leading from cause to effect, and if cause is removed then the effect is less likely to occur, Cetris Peribus.
Resources for Question 4: - Sandy Duong's Teach One Slides► YouTube►
- Markov chains explained visually Webâ–º
- Definition of Markov Blanket Wiki►
- Teach One by Chelsea Answerâ–º
- What symptoms of COVID are neurological? What does ChatGPT say are neurological symptoms of COVID?
- What symptoms of COVID are gastrointestinal in nature? What symptoms are inflammatory? What are respiratory symptoms of COVID? PubMed Review►
- Are all of the variables in the network correlated with each other?
- Does knowing that a female patient has a positive test result increase the chance that she is over 30 years old?
- Is it correct to say that patients with cough and fever are likely to have an elevated chance of positive COVID test?
- List the 5 symptoms of COVID that appear latest in progression of the disease.
- Is muscle ache associated with positive COVID test result?
- Is throwing up associated with positive COVID test results?
- What is the Markov Blanket of fever?
- Under what condition is fever associated with chest pain?
- What are the co-parents of fever?
- Is muscle ache cause of fever or the reverse?
- Is sore throat a common cause or a common effect?
- If a patient does not currently have fever, what factors can be used to establish whether the patient will develop fever later?
- What are the backdoor path from COVID test results to loss of smell?
- Is runny nose associated with wheezing? with sore throat?
- Does presentation of COVID depend on the age of the patient?
- What variables are indirect predictors of COVID?
- What is the symptom most involved in mediating impact of indirect variables on COVID test results?
- How many variables are direct predictors of COVID test results?
- Use ChatGPT to identify 10 alternative ways of indicating sore throat? Can a language model be used to predict positive COVID test results?
- Write an equation for calculating probability of fever from patient's age, gender and presentation.
- Write an equation for calculating probability of excessive sweating from patient's age, gender, and presentation
- If the patient reports some symptoms but not others should we assume that symptoms not mentioned are not present?
Resources for answering Question 5: - See Exhibit 20.1 for definition of various terms in Big Data in Health Care, Chapter 20, page 492
## MoreFor additional information (not part of the required reading), please see the following links: - Introduction to causal inference Read 1► Read 2► Video► Slides►
- Meta analysis through Bayesian networks Read►
- Introduction to Bayesian networks Read►
- Learning Bayesian Networks Read►
- Selection of Judea Pearl's articles PubMed►
- Applications of Bayesian networks in healthcare PubMed►
- Use of graphs in removing confounding Read►
- Bayesian networks in neuroscience Read►
- Cost analysis using Bayesian networks Read►
- Bayesian network classifiers Read►
- Introduction to Markov process Tim's Lecture►
- Explanation of predictions Aloudah's Lecture►
This page is part of the course on Comparative Effectiveness by Farrokh Alemi, Ph.D. Course Home► Email► |