What is PLS Predict and how to report it?
Вставка
- Опубліковано 14 лип 2024
- The PLS predict algorithm has been developed by Shmueli et al. (2016). The method uses training and holdout samples to generate and evaluate predictions from PLS path model estimations.
The research by Shmueli et al. (2016) proposes a set of procedures for prediction with PLS path models and the evaluation of their predictive performance. These procedures are combined in the PLSpredict package github.com/ISS-Analytics/pls-... for the statistical software R. They allow generating different out-of-sample and in-sample predictions (e.g., case-wise and average predictions), which facilitate the evaluation of the predictive performance when analyzing new data (that was not used to estimate the PLS path model). The analysis serves as a diagnostic for possible overfitting of the PLS path model to the training data.
In this webinar, r. Soumya Ray talks about what is predition? How is Predcition done in PLS? What are the current reporting standards for PLS Predict ? What are its limitations and how to deal with them?
It is very informative and new insights on PLS, particularly on PLS predict. Thank you, Dr. Faizan and Dr. Ray!
MP Low I appreciate your comment. Please recommend the channel within your circles
This was a very helpful discussion. Thank you!
That's an amazing effort.... All the best. For sure I'm gonna share this knowledge with my circle.
Thank you very much. I will really appreciate it
Wow, amazing! Regards from Brazil!
Thank you for liking the content. Please share it further.
amazing lecture. Thank you sir!
Thank you for your comment. Please help me by sharing it further
@@ResearchBeast Sure sir!
The session is really really informative, detailed and they are easy to understand. You have turned technicalities into practical! alhamdulillah terima kasih!
Thank you so much. Please help me sharing it further.
Sure. No wonder Prof Soumya is one of the pioneers of PLSpredict@@ResearchBeast
Thank you very much. Your video has given me a lot of insights on how I should approach the discussion of my paper. Everything about PLS is new to me and it is my first time to use it. I have been focusing so much on the tool's "metrics" but Dr. Ray's message helped me to get back on track - tell great and interesting stories from the predictive view of my data.
Thank you. Glad that you like the video. Please help by sharing it further
Thanks a lot 👌👏
Simo Veh my pleasure. If you like this type of content, please share the channel within your circles.
@@ResearchBeast of course thank you again
Amazing lecture and thank you so much!
I have a question regarding the k (number of folds)value when running the SmartPLS predict procedure. The default value of k is 10.
But in my case, I have 62 samples. When evaluating the out-of-sample prediction if I accept the k (fold) =10, 62 can not divide 10 into a whole number, it is a decimal number 6.2. So it means that I will have 6.2 samples in each subgroup. That sounds weird maybe.
So for my 62 samples, do you have any suggestions for my k (number of folds) value that I should set?
Looking forward to your reply! Thanks in advance.
Thank you
You are welcome. Please help me sharing it further
This has been very informative for me. Dr please can we get access to the powerpoint presentation?
what if all PLS SEM RMSE more than LM RMSE?
How this PLS Predict is different then the usage of ANN. Many recent papers are applying Hybrid PLSSem-ANN. Please elaborate. Thanks
For plspredict, One of my endogenous constructs PLS, Q2 below than the LM, Q2. All indicators, Q2 value under that particular construct below LM, Q2. however another construct values achieved the recommended threshold. Pls advice if thats acceptable?