ICACT20210173 Question.1
Questioner: ahmed.ismail@e-campus.uab.cat 2021-02-09 오후 3:38:50 |
ICACT20210173 Answer.1
Answer by Auhor n.brian20202@gmail.com 2021-02-09 오후 3:38:50
Chrome  Click!! |
| In your paper, is there a relation between the number of access points and the accuracy of your proposed method ? |
Thank you for your question.
Since the number of access points is input, we believe that the number of access points and the accuracy cannot be separated.
In the future, I think it will be necessary to check the accuracy when the number of access points is reduced. |
ICACT20210173 Question.2
Questioner: js.kim@etri.re.kr 2021-02-09 오후 3:45:36 |
ICACT20210173 Answer.2
Answer by Auhor n.brian20202@gmail.com 2021-02-09 오후 3:45:36
Chrome Click!! |
| In this paper, the classification model and the esemble learning showed performance lower than the conventional methods. I wonder the reason why the classification model and the esemble learning are worse than the conventional methods. |
Thank you for your question.
I'm using deep learning, so I'm not sure why I'm printing this result.
When I checked the results while referring to the data myself, if the data became abnormal for some reason, it was thought that the cause was that the model could not repair it and the output values were far apart. |
ICACT20210173 Question.3
Questioner: js.kim@etri.re.kr 2021-02-09 오후 4:18:21 |
ICACT20210173 Answer.3
Answer by Auhor n.brian20202@gmail.com 2021-02-09 오후 4:18:21
Chrome Click!! |
| The result of accuracy comparison in Figure 4 is not clear. Is the regression analysis really superior to other methods? |
Thank you for your question.
Regression analysis is shown by the red line, but compared to other methods, the closer the error is to 0 [m], the larger the number, and the larger the error, the smaller the number, so I think the accuracy is good. |
ICACT20210173 Question.4
Questioner: 524078785@qq.com 2021-02-09 오후 4:12:53 |
ICACT20210173 Answer.4
Answer by Auhor n.brian20202@gmail.com 2021-02-09 오후 4:12:53
Chrome Click!! |
| How does the accuracy of the positioning method used by the author compare with the Fingerprint method? |
Thank you for your question.
Since the classification model is a learning method similar to the Fingerprint method, no comparison is made.
However, the result is not so good, so I think it will be necessary to compare it with the Fingerprint method in the future. |
ICACT20210173 Question.5
Questioner: 17862328070@163.com 2021-02-09 오후 3:56:48 |
ICACT20210173 Answer.5
Answer by Auhor n.brian20202@gmail.com 2021-02-09 오후 3:56:48
Chrome Click!! |
| Which models do you use in the paper? What is the input to the model? what is the size of data set? thank you! |
Thank you for your question.
The models we are using are a classification model and a regression analysis model.
The input is the signal strength received from the access point.
The size of the dataset is about 2000. |
ICACT20210173 Question.6
Questioner: js.kim@etri.re.kr 2021-02-11 오후 6:17:05 |
ICACT20210173 Answer.6
Answer by Auhor n.brian20202@gmail.com 2021-02-11 오후 6:17:05
Chrome Click!! |
| What is the future work? |
Thank you for your question.
From now on, further improve the accuracy by changing the parameters of the regression analysis model. Also, collect data at a different location than this time and check the accuracy. |
ICACT20210173 Question.7
Questioner: rusly.eic04g@nctu.edu.tw 2021-02-11 오후 6:21:20 |
ICACT20210173 Answer.7
Answer by Auhor n.brian20202@gmail.com 2021-02-11 오후 6:21:20
Chrome Click!! |
| Thank you for the presentation. If there is a chance to reconfigure distributed sensor network setup, by increasing(or decreasing) number of nodes is there a chance to do it on-the-fly or you need to retrain whole network? |
Thank you for your question.
Since the content of the question is not considered this time, I will answer it on the assumption.
If you want to rebuild your network setup, you probably need to retrain.
If the number of nodes hasn't changed, you may not need to retrain, but I think it's better to train to ensure accuracy. |
