Limits...
Optimal Sensor Selection for Classifying a Set of Ginsengs Using Metal-Oxide Sensors.

Miao J, Zhang T, Wang Y, Li G - Sensors (Basel) (2015)

Bottom Line: The relation of the minimum numbers of sensors with number of samples in the sample set was revealed.The results showed that as the number of samples increased, the average minimum number of sensors increased, while the increment decreased gradually and the average optimal classification rate decreased gradually.Moreover, a new approach of sensor selection was proposed to estimate and compare the effective information capacity of each sensor.

View Article: PubMed Central - PubMed

Affiliation: State Key Laboratory of Industrial Control Technology, Institute of Cyber Systems and Control, Zhejiang University, Hangzhou 310027, Zhejiang, China. jiacheng@zju.edu.cn.

ABSTRACT
The sensor selection problem was investigated for the application of classification of a set of ginsengs using a metal-oxide sensor-based homemade electronic nose with linear discriminant analysis. Samples (315) were measured for nine kinds of ginsengs using 12 sensors. We investigated the classification performances of combinations of 12 sensors for the overall discrimination of combinations of nine ginsengs. The minimum numbers of sensors for discriminating each sample set to obtain an optimal classification performance were defined. The relation of the minimum numbers of sensors with number of samples in the sample set was revealed. The results showed that as the number of samples increased, the average minimum number of sensors increased, while the increment decreased gradually and the average optimal classification rate decreased gradually. Moreover, a new approach of sensor selection was proposed to estimate and compare the effective information capacity of each sensor.

No MeSH data available.


(a) Classification performance of sample set A1 with N (1 to 12) sensors; (b) Corresponding TOP 10 and AVERAGE value; (c) Classification performance of sample set A2 with N (1 to 12) sensors; (d) Corresponding TOP 10 and AVERAGE value.
© Copyright Policy
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC4541866&req=5

sensors-15-16027-f004: (a) Classification performance of sample set A1 with N (1 to 12) sensors; (b) Corresponding TOP 10 and AVERAGE value; (c) Classification performance of sample set A2 with N (1 to 12) sensors; (d) Corresponding TOP 10 and AVERAGE value.

Mentions: For a certain sample set A,the average top 10 classification performance of sample sets with N sensors represents the optimal result with sensor set of N sensors. As N increases, the maximum ‘optimal result’ will be obtained with a certain value of N and this value defines the minimum number of sensors needed for discriminating sample set A. Hence, firstly, the overall classification performance of 4095 sensor sets for 502 sample sets were calculated. Due to limited space, the classification performances of sample set A1 = 1, 2, 3, 4, 5, 6, 7, 8, 9 with all nine classes of sample and A2 = 2, 5, 9 with Nos. 2, 5, 9 samples are shown as typical examples in Figure 4a,c. A1 is the most complex classification target in our experiment; A2 was randomly selected with fewer kinds of samples and considered to be less complex compared to A1. For a certain sample set, there areclassification performances with N sensor sets (N = 1 to 12) sensors. The average of the top 10 classification performances (TOP 10) and the average of all classification performances (AVERAGE) with N sensors are shown in Figure 4b,d for sample set A1 and A2 (when N = 12, TOP 10 = AVERAGE, for there is only one sensor set of 12 sensors). The number of sensors where TOP 10 achieves a maximum value is defined as the minimum number Nmin (A) for the discrimination of sample set A, and if TOP 10 has more than one maximum value, the minimum N with maximum value is taken as Nmin (A).


Optimal Sensor Selection for Classifying a Set of Ginsengs Using Metal-Oxide Sensors.

Miao J, Zhang T, Wang Y, Li G - Sensors (Basel) (2015)

(a) Classification performance of sample set A1 with N (1 to 12) sensors; (b) Corresponding TOP 10 and AVERAGE value; (c) Classification performance of sample set A2 with N (1 to 12) sensors; (d) Corresponding TOP 10 and AVERAGE value.
© Copyright Policy
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC4541866&req=5

sensors-15-16027-f004: (a) Classification performance of sample set A1 with N (1 to 12) sensors; (b) Corresponding TOP 10 and AVERAGE value; (c) Classification performance of sample set A2 with N (1 to 12) sensors; (d) Corresponding TOP 10 and AVERAGE value.
Mentions: For a certain sample set A,the average top 10 classification performance of sample sets with N sensors represents the optimal result with sensor set of N sensors. As N increases, the maximum ‘optimal result’ will be obtained with a certain value of N and this value defines the minimum number of sensors needed for discriminating sample set A. Hence, firstly, the overall classification performance of 4095 sensor sets for 502 sample sets were calculated. Due to limited space, the classification performances of sample set A1 = 1, 2, 3, 4, 5, 6, 7, 8, 9 with all nine classes of sample and A2 = 2, 5, 9 with Nos. 2, 5, 9 samples are shown as typical examples in Figure 4a,c. A1 is the most complex classification target in our experiment; A2 was randomly selected with fewer kinds of samples and considered to be less complex compared to A1. For a certain sample set, there areclassification performances with N sensor sets (N = 1 to 12) sensors. The average of the top 10 classification performances (TOP 10) and the average of all classification performances (AVERAGE) with N sensors are shown in Figure 4b,d for sample set A1 and A2 (when N = 12, TOP 10 = AVERAGE, for there is only one sensor set of 12 sensors). The number of sensors where TOP 10 achieves a maximum value is defined as the minimum number Nmin (A) for the discrimination of sample set A, and if TOP 10 has more than one maximum value, the minimum N with maximum value is taken as Nmin (A).

Bottom Line: The relation of the minimum numbers of sensors with number of samples in the sample set was revealed.The results showed that as the number of samples increased, the average minimum number of sensors increased, while the increment decreased gradually and the average optimal classification rate decreased gradually.Moreover, a new approach of sensor selection was proposed to estimate and compare the effective information capacity of each sensor.

View Article: PubMed Central - PubMed

Affiliation: State Key Laboratory of Industrial Control Technology, Institute of Cyber Systems and Control, Zhejiang University, Hangzhou 310027, Zhejiang, China. jiacheng@zju.edu.cn.

ABSTRACT
The sensor selection problem was investigated for the application of classification of a set of ginsengs using a metal-oxide sensor-based homemade electronic nose with linear discriminant analysis. Samples (315) were measured for nine kinds of ginsengs using 12 sensors. We investigated the classification performances of combinations of 12 sensors for the overall discrimination of combinations of nine ginsengs. The minimum numbers of sensors for discriminating each sample set to obtain an optimal classification performance were defined. The relation of the minimum numbers of sensors with number of samples in the sample set was revealed. The results showed that as the number of samples increased, the average minimum number of sensors increased, while the increment decreased gradually and the average optimal classification rate decreased gradually. Moreover, a new approach of sensor selection was proposed to estimate and compare the effective information capacity of each sensor.

No MeSH data available.