Major Histocompatapility Complex

The major histocompatibility complex (MHC) glycoprotein family is an important component of the immune system. In humans, these molecules are also known as human leukocyte antigens (HLA). Each individual has multiple highly polymorphic isoforms of MHC family genes. Classically studied MHC isoforms have been identified as HLA - A, B, C (class I) and HLA - DR, DQ, DP (class II). These molecules play a central role in the central function of the immune system, distinguish themselves from non-self, which is the basis for acquiring immunity against pathogenic microorganisms and diseases, and are critical for organ and tissue transplantation It represents a failure.

Imagine there is a record of C's major code on the piano. That sound is a combination of three sounds C, E, G, and it is gathered up into a complicated sound. In order to find C, E, G I want to break down complex sounds into individual sounds. This is exactly the same way of thinking. We use a neural network with memory that affects recurrent neural networks - future predictions. This is because each predicted character affects the likelihood of predicting the next character. For example, if you say "HEL" so far, you may say "LO" after "Hello". We rarely say anything we can not speak like "XYZ". Thus, the memory of previous predictions helps the neural network make more accurate predictions.

Regularization punishes neural networks to use complex architecture. The complexity of this method is measured by the weight of the neural network. Regularization is accomplished by adding terms to the weighted squared error objective function. This depends on the magnitude of the weight. This is equivalent to adding a prefix. This basically makes the neural network believe that the approximation function is smooth. Where n is the number of weights in the neural network. The parameters α and β control the extent to which the neural network affects the data. Bayesian analysis and optimization can be used to derive appropriate values ​​for alpha and beta. In this wonderful chapter please explain this and the above in more detail.