Machine learning is a powerful way to improve the traditional experiment process. I learned in this seminar about how to use active learning in building models can evaluate data in real time. This method can save time in data collection as well as in the result characterization. In this seminar, the speaker mentioned the potential of this technique can have efficiency with analysis data.
What I have learned from the seminar:
●Process data including data selection and correction.
●How to build a useful model.
●The important of creating a deep learning model for active learning process.
The most useful information that I got from this talk is that giving a goal for the machine to achieve was not a simple task in the past, but now we have better, faster, stronger machine learning tools to help us distinguish and controlling data collection in the simulation model. One question I want to ask is what if the correction data was not going the right way, do you need to rebuild the model?
留言列表
