This paper focuses on adaptive soft sensor modeling based on an improved just-in-time learning and random mapping PLS for chemical processes. The application is relevant. Here are some minor suggestions for it.
1. The English of the manuscript should be improved before resubmission. Some sentences do not read well.
2. Many of the references are not updated. The authors should include more research work on soft sensor within the last 3 years.
3. The main contributions are not clear for this paper. It is necessary to summarize them in the introduction.
4. It is better to give a flowchart of the developed method.
5. How to determine the parameters of the developed method, like the number of local samples, the adjustable parameter, etc? Some parameter sensitivity can be added.
6. There are very recent deep learning techniques for soft sensor. It is better to introduce related works in the introduction, like Quality prediction modeling for industrial processes using multiscale attention-based convolutional neural network, Variable correlation analysis-based convolutional neural network for far topological feature extraction and industrial predictive modeling, Attention-based interval aided networks for data modeling of heterogeneous sampling sequences with missing values in process industry, etc.