用学术英语来翻译下面这段话：迁移学习Transfer Learning的基本思想是利用已解决问题的策略去解决待解决的问题也就是把现有的经验迁移过去13。目前是作为机器学习下的一个分支大多使用神经网络的方法以训练模型为主了通常就是那一套模型的参数去做另一套的初始值减少模型训练的复杂度。但迁移学习提出的初衷不是为了模型训练而是节省人工标注样本的时间让模型可以通过已有的标记数据source domain

The fundamental concept of Transfer Learning is to apply the strategies that have been used to solve one problem to solve another problem, i.e. transferring the existing experience [13]. Currently, it is a branch of Machine Learning mostly using neural network methods for model training, wherein the parameters of one set of models are used as the initial values for another set to reduce the complexity of model training. However, the original intention of Transfer Learning is not for model training, but to save the time required for manual annotation of samples, allowing the model to transfer from labeled data (source domain data) to unlabeled data (target domain data), thus reducing the demand for sample size in Machine Learning. Generally, Transfer Learning involves multiple source domains for one target domain or a single source domain for one target domain, with the latter being the most common case represented by equation (5)