It is an interesting study that tackles one of very important questions in computational neuroscience - how generalisation across stimuli and environments is achieved. It is similar to the concept of schemas, which are thought to primarily rely on frontal cortical areas. In this particular case the focus is on entorhinal grid cells and hippocampal place cells, which authors assert code for the environment and conjunction of environment and stimulus respectively. The authors present a computational model that aims to address the question of whether place cells encode a conjunctive outcome of environment and stimulus representations. It is an interesting hypothesis, which if shown convincingly, would be a major breakthrough in neuroscience.

The k-nearest neighbor algorithm is applied to different classification and regression problems. The closest k training samples are used to predict the class of new input data, i.e., the most similar samples already known are used to classify an unknown data sample. Since the sci-kit library provides all the necessary tools to work on this algorithm, you can use these 5 steps to build your own KNN classifier in Python! As usual, start with importing all necessary libraries needed. This command builds an easy to handle data frame and decreases the complexity of working on the data set.