Inner Speech Classification Project This project focuses on classifying inner speech using advanced machine learning models. By leveraging Transformer Neural Networks (TNN), Long Short-Term Memory (LSTM) networks, and t-distributed Stochastic Neighbor Embedding (t-SNE), we aim to capture the temporal dynamics and high-dimensional structure of inner speech data for effective classification. The project uses implementations of the UMAP algorithm (Uniform Manifold Approximation and Projection) for feature selection which is demonstrated in the processing files. Additional tested models with lower accuracy are recorded as well.