Hello everyone,
I am conducting a study in which I pretend to relate work events with emotions and I'm doing so with neural nets. So, as far as I know, I can interpret each hidden node as a "category" (or a factor) where related events group together before generating outputs. It's similar to a classification system, but resorting to a non-linear method. Now, I searched and searched for previous studies that did something similar, regarding classification of events in organizational research and classified them with neural nets but apparently there are none. Does anyone know of enlighting researches that followed this same method?
TIA
See:
Fault Diagnosis of Nonlinear Systems Using a Hybrid Approach, Sobhani-Tehrani, Ehsan, Khorasani, Khashayar, Springer, 2009.
Computational Intelligence in Fault Diagnosis
Palade, Vasile, Bocaniala, Cosmin Danut (Eds.), Springer, 2006.
As emotions have a different meaning acording to your background, you will need to establish some restrictions on your research. For instance, you can focus on fear with several levels of fears, in the same way you can focus on happiness with several levels of happiness. Having an ample set of emotions can make the intrepretations of your results difficult. It seems that in your example the ANN inputs could be a set of variables describing work events, while each ANN output would be one emotion (fear, happiness, etc.) I worked on emotions and learning https://dl.dropboxusercontent.com/u/74744282/Q10_26298.pdf.
This is was one of the paper i referred when i was working on computational social science. hope this will help.
Topic: Emotions and Organizational Behaviour
Hi all,
Rita, this paper could be interesting for you:
http://www.tandfonline.com/doi/abs/10.1080/13594320902995916#.VHYDijGG-VM
Depending on your knowledge of NN at this moment could be useful for you.
Best,
Pedro
- Badges
- Science topic