In thin-walled milling experiments, I collected the whole process of cutting force signal, due to my final research is to use the type value points of the information on the thin wall piece, so I have to type cutting force signal and the value information match point, so I should be on the cutting force signal is split, excuse me I how to split the cutting force signal, in order to ensure better corresponding relation.In addition, since I subsequently adopted a deep learning model, I need to normalize the data in advance. I would like to ask whether I should first normalize the overall cutting force signal and then split it or normalize each segment after splitting it.
For the 100 measuring points in the figure, add the cutting force values of 1 million moments in the processing process that I have collected, and how can I match the values of measuring points (such as deformation) with the cutting force at this point?
Yes, data normalization, data down sizing., e.g., via PCA/LDA/ICA etc., are first data pre-processing steps. You then have to generate groups, or clusters of data and then data splitting for training and testing or validation purposes.
I would normalise the data at the input of each stage, with a denormalisation at the output of each stage. If you do not denormalise, errors may be biased.
I would take an approach similar to the one in statistics, if the different variables you are inputting in your model have a considerable difference in scale I would recommend to normalize it. If the scale is naturally close in the different input variables I would start by not normalizing and check the accuracy of the model.
How everybody said normalization is a preprocessing step. I recommend you normalize your data prior to make some operation. With this, you will obtain a low computational time and you will reduce variation problems. This step is recommended when the features of your dataset are in different ranges. Good luck!!
José Alberto Guzmán Torres Diego Andrade Remy Chen Uma Mahesh R N Will Serrano
Thanks for everyone's answer. Through your answer, I understand that the normalization operation is necessary and important.I also know some other methods of data preprocessing from his answer.To summarize, are you suggesting that I do normalization first and then split the data?
feature normalization can also be done and when the model overfits on data. data augmentation can be applied to reduce the overfitting situation.
It all depends on the algorithm you're employing. Ranges, for example, are irrelevant in Random Forests because one feature is never compared to another in terms of magnitude. At each point, only one feature's range is split.
However, SVM or Logistic regression would likely perform better if the features are approximately the same magnitude, unless you know a priori that one feature is significantly more critical than others, in which case a larger magnitude is appropriate.
Well Teng Zhang, in summary, I recommend you first split the data and then apply the normalization process. Good luck!!
The main objective of normalization is to change the values of numeric columns in the dataset to a common scale, without misrepresenting differences in the ranges of values.
However, in ML it is not all datasets that require normalization, it is only essential when dataset features have different ranges that can lead to overfitting or underfitting problems and subsequently results to bad performance.
Check these it might be helpful
https://www.analyticsvidhya.com/blog/2020/04/feature-scaling-machine-learning-normalization-standardization/
https://www.kdnuggets.com/2020/04/data-transformation-standardization-normalization.html
This is an interesting and important question because normalization can improve the performance of the developed machine learning models considerably. Please have a look at the below reference:
https://towardsai.net/p/data-science/how-when-and-why-should-you-normalize-standardize-rescale-your-data-3f083def38ff
The rudimentary purpose of the normalization of the dataset is to measure the values of the dataset in a common scale if you have a dataset with different ranges samples. Because normalization can help you to reduce the overfitting problems that can make the loss lower as well with a good accuracy. 😊
of effectively organizing data in a database. There are two goals of the normalization process: removing redundant data (for example, storing the same data in more than one table) and ensuring that data dependencies make sense (only storing related data in a table).
- Badges
- Science topic