Automatic Stress Detection in Working Environments From Smartphones’ Accelerometer Data: A First Step
Citations
9,185 citations
255 citations
201 citations
186 citations
References
45,034 citations
"Automatic Stress Detection in Worki..." refers background in this paper
...(5) [33], Energy: which is the sum of the squared discrete FFT component magnitudes of the signal....
[...]
[...]
38,291 citations
"Automatic Stress Detection in Worki..." refers methods in this paper
...In order to check the effect size of each of the features we computed the Cohen’s d effect size and quantified it using the thresholds defined in [36], i.e., |d| 0.2 ‘negligible’, |d| 0.5 ‘small’, |d| 0.8 ‘medium’, otherwise ‘large’....
[...]
20,196 citations
14,144 citations
"Automatic Stress Detection in Worki..." refers background in this paper
...The Random classifier had an ACC1 = 0.81. this is because an output will be counted as an error only if the prediction is low and the actual class is high or vice versa....
[...]
9,185 citations
"Automatic Stress Detection in Worki..." refers methods in this paper
...For this purpose, we are going to use two classification models namely: 1) Naive Bayes [37] (pp....
[...]
...With respect to the general model, the similar-users model had an increase of 8% in accuracy for Naive Bayes and 5% for the Decision Tree....
[...]
...The Ordinal Naive Bayes was omitted since it did not present any performance improvement in the previous cases....
[...]
...Table V shows the results for the similar-users model for Naive Bayes and Decision tree....
[...]
...The Ordinal Naive Bayes classifier did not present any improvement over the traditional Naive Bayes....
[...]