This project explores the application of deep learning and machine learning techniques to enhance psychological testing. Centered on investigating the feasibility of reducing the number of questions in a psychological test without compromising the validity of the results.
The conclusions of this projects aim to answer the following reflection: it is preferable to maintain the number of original questions with ideally more valid results or to reduce the number of questions so that the user is more likely to take the test and maintain attention during the test despite theoretically not having full accuracy.
To answer this, deep learning and machine learning techniques are employed. The methodology includes researching the theory and basis of the most commonly used psychological tests, training different models with data open source data from a psychological test, using different feature selection techniques to identify the most relevant questions, and then training optimized models with a reduced number of features to compare the outcomes.
For more information, results, references and description of the project, please refer to the paper published in this repository.