############################################### --Contact Details-- Name: Kristian Callaghan Institution: University of Bath, Dept. Electronic and Electrical Engineering Email: KMSC20@bath.ac.uk Date: October 2024 ############################################### --Overview-- Dataset containing all Volatile Organic Compound sensor data and ripeness stage labels for tests 3 & 4 only. ############################################### --Contents-- /VOC_x_test.csv: CSV file containing cleaned data from the Figaro TGS2620 & TGS2602, and Sensiron SGP40 sensors for tests 3 & 4 for ML testing purposes. This file contains 30% of all recorded data and is class balanced. A header row is included. The first column is the index column and does not have a header. Each row represents an instantaneous reading of the Figaro TGS2620 & TGS2602, and Sensiron SGP40 sensors at a point in time and is given a unique index number in the index column. The index number corresponds to the labels in VOC_y_test.csv. The header row titles are defined at the bottom of this ReadMe. /VOC_x_train.csv: CSV file containing cleaned data from the Figaro TGS2620 & TGS2602, and Sensiron SGP40 sensors for tests 3 & 4 for ML training purposes. This file contains the 70% of all recorded data NOT present in VOC_x_test.csv, and is class balanced. A header row is included. The first column is the index column and does not have a header. Each row represents an instantaneous reading of the Figaro TGS2620 & TGS2602, and Sensiron SGP40 sensors at a point in time and is given a unique index number in the index column. The index number corresponds to the labels in VOC_y_train.csv. The header row titles are defined at the bottom of this ReadMe. /VOC_y_test.csv: CSV file containing the labels for data from tests 3 & 4 for ML testing purposes. The labels in this file correspond to the data in VOC_x_test.csv and are class balanced. Two columns are included. An unlabelled index column and the label column, where the labels are integer values from 1 to 5 corresponding to ripeness stages for each instantaneous sensor reading. Please refer to the associated publication for explanations of the ripeness stages and how they are derived. /VOC_y_train.csv: CSV file containing the labels for data from tests 3 & 4 for ML training purposes. The labels in this file correspond to the data in VOC_x_train.csv and are class balanced. Two columns are included. An unlabelled index column and the label column, where the labels are integer values from 1 to 5 corresponding to ripeness stages for each instantaneous sensor reading. Please refer to the associated publication for explanations of the ripeness stages and how they are derived. ***** Header Row Definitions ***** "TGS20" = Internal chamber VOC reading from a Figaro TGS2620 chip. "TGS02" = Internal chamber VOC reading from a Figaro TGS2602 chip. "SGP" = Internal chamber VOC reading from a Sensiron SGP40 chip. ############################################### --Citation-- ###############################################