Merge branch 'PE_RC_2023' into 'master'

Updated to june 2023

See merge request !2

slides/Sven/HPC.Rmd

 ---
 title: "High performance computing with R"
 author: "Sven Warris"
-date: "November 2021"
+date: "June 2023"
 output: beamer_presentation
 ---
 
@@ -122,7 +122,7 @@ Cons:
 
 # GPGPU performance
 
-![performance](./images/floating-point-operations-per-second.png)
+![performance](./images/floating-point-operations-per-second.png){height=80%}
 
 # HPC Overall message
 
@@ -240,7 +240,7 @@ clusterExport(cl, c("isNatural", "isIntpower"),
               envir=environment())
 ```
 
-Exercise: check number of cores on our server. Should be more than one...
+Exercise: check number of cores on your laptop or server.
 
 # Apply
 
@@ -447,11 +447,11 @@ c(12,41,53,8,9,2,11,7)
 
 # Serial vs. Parallel
 
-![SerialParallel](./images/serialParallelProblem.png)
+![SerialParallel](./images/serialParallelProblem.png){height=80%}
 
 # Problem solving: Domain decomposition
 
-![Domain](./images/domainDistributions.png)
+![Domain](./images/domainDistributions.png){height=80%}
 
 # Problem solving: Functional decomposition
 

slides/Sven/machineLearning.Rmd

 ---
 title: "Big data and machine learning"
 author: "Sven Warris & Christina Papastolopoulou"
-date: "November 2021"
+date: "June 2023"
 output: beamer_presentation
 ---
 
@@ -113,6 +113,15 @@ Machine learning can be used for this:
     - Training / testing difficult
     - Data biases
     - Clustering, neural networks
+
+## ChatGPT and large language models
+
+- language model consisting of a neural network with many parameters
+- trained on large quantities of unlabeled text
+- self-supervised learning or semi-supervised learning.
+- ChatGPT
+    - extremely helpful for sorting out programming issues
+    - prompt hacking: asking questions is key
     
 ## General setup of a ML experiment
 
@@ -307,7 +316,7 @@ Use parallel processing to train, test and evaluate many different random forest
 
 ## Putting it all together
 
-Create a machine learning report in R markdown:
+Create a machine learning report in R markdown or notebook:
 
 - Data set:
     - Capsicum data
@@ -320,11 +329,9 @@ Create a machine learning report in R markdown:
 - Create decision tree
 - Create random forest models
 - Report on the most accurate model
-- Send me your report (sven.warris@wur.nl)
+
 
 ## Round up
 
 That's it! Thank you for your attention!
 
-- Ron Wehrens: ron.wehrens@wur.nl
-- Sven Warris: sven.warris@wur.nl

slides/Sven/reproResearch.Rmd

 ---
 title: "Reproducible Research"
 author: "Sven Warris & Maikel Verouden"
-date: "November 2021"
+date: "June 2023"
 output: beamer_presentation
 ---
 
@@ -49,7 +49,7 @@ The "what" that needs to be reproduced is typically:
 * reproducibility makes it easier to check the research
 * it is easier to reproduce the research independently
 * easier to extend the research
-* reusable code and instruction resulting in increased efficiencies
+* reusable code and instruction resulting in increased efficiency
 
 # Exercise
 
@@ -72,7 +72,7 @@ Examples:
 * Co-develop
 * Develop-test-release cycle support
 * Many, many tools available
-* Integrated into R=
+* Integrated into R-studio 
 * [WUR Git lab](https://git.wur.nl)
 * [github](http://www.github.com)
     * Allows for **DOI**!
@@ -193,6 +193,7 @@ In the **Markdown Quick Reference** you will also find a section on **_Inline R
 
 - Yihui Xie Chunk Options : https://goo.gl/Bg1c5G
 
+
 # Render your document
 
 Once your document is finished, it is time to render the document in order to create the final output.