Introducing Artificial Neural Networks in SPSS | Analysis and Interpretation
HTML-код
- Опубликовано: 23 авг 2024
- In this video, I will demonstrate how to run an artificial neural network (ANN) analysis in SPSS and interpret the output. The concepts discussed include multilevel perceptron ANN, deep learning, hidden layer, accuracy, specificity, sensitivity, training, and testing.
Related papers and books are as follows:
1. Cambridge-Michigan report: michiganassess...
2. Research paper: www.tandfonlin...
3. Quantitative book: www.routledge....
![Physics Informed Neural Networks (PINNs) [Physics Informed Machine Learning]](/img/1.gif)
Thanks, best explanation!
So how can we build the model out of the whole Ann’s analysis? Is there any model as an output?
thanks, and now I'm an expert in SPSS ANN analysis! haha
Thank you, prof. for this amazing session. I faced an issue, could you please tell me why in the output the network performance is not activated. it gives me a model summary only?. I will be very grateful for your reply. Thank you
Thank you so much for this Useful presentation, Please could you provide me with information on how could we interpret the output, thanks in advance.
Hi, for more information, please see this: michiganassessment.org/wp-content/uploads/2020/02/20.02.pdf.Res_.PredictingListeningItemDifficultywithLanguageComplexityMeasures-AComparativeDataMiningStudy.pdf
Thanks for the video. I have some questions about this.
1. The "Classification" table does not appear in my SPSS calculations.
2. I have seen R²(R² = 1-RMSE/S², where S² is the variance of the test data’s desired output) calculated in some papers, but i don't know exactly how it is calculated.
Hoping for a response
Thanks
Great video. I want to make forest fire susceptibility mapping using ANN in SPSS. but i don't know how to go about it...please help me
Hello thank You for this video, i want to perform Neural network in SPSS and generate data such as roc curve to prove my model that i did in regression, my question is my data are continuous variable (BMI) , but i feel only nominal variable are acceptable to be dependent in ANN , how can i change my data … thank You
Hello mate! Im using spss 26, however under output the classification results and Roc curve is greyer out. Any help?
While applying ANN the output tab does not allow me to select "Classification result" "ROC Curve" extra do you have any suggestions for it.
Good tutorial. Just a question, how and where is the equation that we can use or apply in actual data that we just input the variables to get the final result ? Thanks.
NNs are black boxes, so you will not output the math equations (since there are many of them, and not just one), unlike regression models where there is only one equation.
@@VahidAryadoust thanks for your reply. Actually i m doing AI in regression and wanted to try out ANN. thanks.
ANN are black boxes. This program does not output the equations.
Could we replicate the same results?
Great presentation! I have a question. I have noticed something quite odd. When I create a neural network I get a certain output. But when i run it again, with exactly the same settings, the output (Importance, Normalized Importance) all change - why? How do i know which model is the best, given that "Percent Incorrect Predictions" always is zero or close to it?
That is to be expected, because you do not fix the 'Seed". Go to the "Seed" option and fix it to the same number every time you run it.
Thank you very much
Hey Vahid. Thank you for sharing this wonderful content. I am just wondering how you were able to click and select the percentages and then draw the bar chart. My SPSS 26 does not let these values select. Thank you!
You might want to click twice or three times quickly. That should do the trick.
@@VahidAryadoust Thanks. I will try and let you know if it works or not.
Thank you for helping via this video.If have a one simple question what are the required variables to apply Neural networks?
Sorry, I am not sure about the question. Can you elaborate what you mean?
@@VahidAryadoust can i apply neural network for discrimination of varieties in the field of food science??
Thankyou for the information. I have a question, can we use the same for ecosystem restoration predictions like predicting the years required for the restoration of degraded site by using values obtained from physicochemical analysis?
Yes, you can.
Can you please tell me how can I run Ann after SPSS Amos.
Hello Vahid, thank you for your presentation.
There is one question that I can not seem to find an answer.
I want to apply my NN model to different cohorts, but I don't know how to do it.
For the logistic regression, I would do it by the "compute" command and write down the regression function.
But I don't know how to apply my NN model to a new cohort.
I found the answer to this problem. I tried a a couple years ago to do it but I could not do it, so the answer to my students was the same as everyone tells you on the net and here in RUclips: Neural Networks are Black Boxes. This is no true. The REAL ANSWER IS: You have to export the Neural Network as XML from the principal menu: Analyze - Neural Network - Multilayer Perceptron - Export (after you did all that you need to do to get the neural network model). So, import a new dataset with the same independent variables but without the depent variable, use the Utilities menu - score wizard (in Spanish is Asistente de puntuación) and you will have to import the XML document to score the dataset. I will do a video in a couple weeks with this steps cause no body in RUclips knows how to do it and everybody say always some crap like "Blackboxes". By the way, I got the answer trying, following steps, giving instructions and making some test with the ChatGPT.
@@christophercuatepotzo5746 You are a genius :D Thank you so very much :D
@@christophercuatepotzo5746 Oh my God. I tried it now and it worked fantastically. Thank you so much :D
Thank you so much for this wonderful explanation. My question is about what would be the relevant data to present in a table for publication. I work in the medical field and I am used to using logistic regression however this alternative is even more powerful. Some of the publications that I have seen include specific numbers for specificity, sensitivity, and precision. Where I could find this data in SPSS? Thank you so much!
specificity, sensitivity are found in the output, as shown in the video.
@@VahidAryadoust thank you, do you refer to the graphic? Where can I find a number instead of the graphic for precision, sensitivity, and specificity? I was not able to find it in the output. I really appreciate your feedback, it’s really helpful
Hi, this is excellent video, thanks a lot. One question, how can we apply this in real life setting. i.e I have data of patients and their clinical profiles and outcomes. Now, if I have a model by this method, so for new patient can I predict the outcome using this method?
Yes, you can. Neural networks can help you make predictions.
Hi, thanks a lot for your videos. What do you think the best alternative to spss is among the free softwares (no programming language)? I don’t know what to choose among jasp, jamovi, gretl. I come from economic background and I’ve covered most important machine learning topics and some deep learning. I wonder if neural nework analysis is available on the softwares I mentioned or any other free software I may not know. I would really appreciate if you can suggest what is in your opinion the best one for advanced analysis.
Jamovi, JASP, and R are pretty fine.
@@VahidAryadoust thank you. I actually downloaded Jamovi, but haven't found some important analyses such as decision tree, KNN, support vector machine ecc. Also no neural network. I've seen that Jasp has it all, which is cool!
@@datascience1274 Yes, JASP provides more ML options.
how can we use it to predict sand storms?
How to replicate the same results please
Hello Vahid, when is it appropriate to use this analysis in investigating study. What are the assumptions of using Artificial Neural Networks in SPSS. It is to test impact, Effect or Influence? I will appreciate your quick response
It is used for prediction. For example give inputs A, B, C... etc., can we predict the amount of an output? The terminology is 'prediction' rather than effect etc. in machine learning.
@@VahidAryadoust Can it be used for nexus between Big Five models, assessment centres and management development. Taking Assessment centres as mediating variable, Big Five is the independent variable while Management Development is Independent variable
@@segunemmanuel2322 I do not have expertise in that field, but as long as you are defining your research problem as a prediction question, ANN can be used as a method.
Vahid Jan, Thanks! Just curious, can these classification methods take the dependence of data points for different subjects into account (i.e., nested data structures like students nested within classroom nested with schools etc.)?
You can create a variable for the nested relationship and enter into the network.
@@VahidAryadoust You mean like mixed-effects modeling (e.g., school_id 1, 1, 1, 1, ... student_id 1, 1, 2, 2, ...)?
@@rezanorouzian8198 Yes, you can use it and also compare the fit of the model with and without this variable.
Can you mâle vidéo in french language ?
is getting different results normal ?
Fix the seed so you won't get different results.
Can you share the data set?
The data cannot be share, I am afraid.
@@VahidAryadoust Good day. For the sake of our academic purposes, may I ask for the data set just for the outline/basis of our concept paper regarding neural network? Thank you so much.
Spss!?
SPSS is a children's (under grad) package. Even at master level you dont use SPSS !
Really ? 😮