1. By reading the background materials, what spatial geometric properties of neurons can you think of as features extracted before classification? (You can check the neuron structure with Neuromantic software) and explain it according to the features extracted by yourself: What is the most important principle of feature extraction? (that is, why choose some features as a reference for identification and classification? )

2. Using the spatial geometric data of the sample neurons in the appendix, according to the geometric characteristics of the five types of neurons mentioned in the subproblem (1) (the intermediate neurons can be further subdivided into three types), an algorithm for spatial morphological classification of neurons is designed. It is necessary to briefly describe the algorithm principle, algorithm flow chart and possible conclusions.

3. The morphology of 3.warp elements is complex and diverse, and the classification and naming of neurons are still the focus of research. The five types of neurons we see are only important categories, and there are actually many other categories. Can you classify and name all neurons according to their geometric characteristics according to your proposed neuron classification method, so as to provide biologists with suggestions on neuron naming? It is required to give naming rules, or name coding methods, or naming principles of symbols and graphics.

4. Over time, the actual morphology of neurons will change, and dendrites and axons will continue to grow. Can you predict the growth and change of neuron morphology? How do these morphological changes affect the geometric features you have determined? And with the growth of neurons, is the naming rule you designed feasible?