Our idea tries to improve purely text-based personality tests by adding other communication cues being assessed with our AI.
This repo contains the code for several neural network trying to predict the outcome of a personality test based on different iput data.
Containes the training data the we created to be able to teach our algotrithms the personality of a canditate.
Simple neural network looking at the most important the we as software develpoes have to make: choosing our favourite editor ot the right programming language for the job etc.
Small application to test the output of nn.py with customazible inputs.
This neural network is trying to find the personality traints of a person based on their image itself. Using the nn_img_test.py script it is possible to run the trained algotimhn on an own image.
This network is filled with data that is collected from humans like the openess of a person to williing a short interview.
| network | loss | accuracy |
|---|---|---|
| nn.py | 0.0345 | 0.7857 |
| nn_img | 0.1383 | 0.4118 |
| nn_human | 0.1273 | 0.4643 |