This contains my code for the final project of the neuromatch academy computational neuroscience programme. This is an encoding GLM for investigating the semantic representation of each hierarchical area in visual cortex. Although the R2 score is not significant, it can be used in the future for education purposes and for improvements upon project. Some recommendations include but not limited to: Changing the encoding method for categorical variables to a more representative one; Increasing the number of independent variables; using a decoding model (Classification instead of regression); trying a different model or regularization penalty term. Deep learning models perform much better than GLM, however, due to higher interpretability of GLM, they make a great tool to infer conclusions about the brain.
Aburas98/NMA-kay-V_cortex_Modelling
This contains my code for the final project of the neuromatch academy computational neuroscience programme. The results are not ideal, but it can be used in the future for education purposes and for improvements upon project.
Jupyter Notebook