Data
Contaminant-detection-in-packaged-cocoa-hazelnut-spread-jars-using-Microwaves-Sensing-and-Machine-Learning-11.0GHz(Urbinati)

Contaminant-detection-in-packaged-cocoa-hazelnut-spread-jars-using-Microwaves-Sensing-and-Machine-Learning-11.0GHz(Urbinati)

active ARFF CC BY-SA Visibility: public Uploaded 31-05-2023 by Luca Urbinati
0 likes downloaded by 0 people , 0 total downloads 0 issues 0 downvotes
Issue #Downvotes for this reason By


Loading wiki
Help us complete this description Edit
Contaminant-detection-in-packaged-cocoa-hazelnut-spread-jars-using-Microwaves-Sensing-and-Machine-Learning-11.0GHz(Urbinati) ---------------- This dataset is part of a series of five different datasets each one measured with a different microwave frequency: 9.0, 9.5, 10.0, 10.5, 11.0 GHz. PAY ATTENTION: THE DATASET PRESENTED IN THIS PAGE HAS BEEN ACQUIRED AT 11.0 GHz! IF YOU ARE LOOKING FOR THE OTHER FOUR DATASETS, VISIT THE OPENML PROFILE OF THE AUTHOR OF THIS DATASET. The following description is valid for all the five datasets. Dataset description. To detect contaminants accidentally included in industrial food, Microwave Sensing (MWS) can be used as a contactless detection method, in particular when the food is already packaged. MWS uses microwaves to illuminate the target object through a set of antennas, records the scattered waves, and uses Machine Learning to predict the presence of contaminants inside the target object. In this application the target object is a cocoa-hazelnut spread jar and each instance (sample) of this dataset consists in 30 scattering parameters of the network composed by: antennas, target object (a jar w/ or w/o a contaminant inside) and medium (i.e. the air) in between. The types of contaminants vary from metal to glass and plastic. Each sample has been measured at five different microwave frequencies that are: 9.0, 9.5, 10.0, 10.5, 11.0 GHz. PAY ATTENTION: THE DATASET PRESENTED IN THIS PAGE HAS BEEN ACQUIRED AT 11.0 GHz! IF YOU ARE LOOKING FOR THE OTHER FOUR DATASETS, VISIT THE OPENML PROFILE OF THE AUTHOR OF THIS DATASET. Data Set Characteristics: :Microwave frequency used for acquisition: 11.0 GHz :Total Number of Instances: 2400 :Total Number of Uncontaminated Instances: 1200 :Total Number of Contaminated Instances: 1200 :Total Number of Classes: 11 :Target: The last column, column 31, contains the class label as integer value :Number of Contaminated Instances Divided Per Class (full explanation and pictures in [2]): - "air_surface" (i.e. cap-shape plastic with the same dielectric constant of the air): 200 - "big_pink_plastic_shere_middle": 100 - "big_pink_plastic_shere_surface": 100 - "glass_fragment_middle": 100 - "glass_fragment_surface": 100 - "small_metal_sphere_middle": 100 - "small_metal_sphere_surface": 100 - "small_plastic_sphere_middle": 100 - "small_plastic_sphere_surface": 100 - "small_triangular_plastic_fragment_surface": 200 "surface" means that the instance was placed on top of the cocoa-hazelnut spread, at the spread-air interface "middle" means that the instance was placed in the middle of the jar filled with cocoa-hazelnut spread :Number of Attributes in a generic Instance: 30 :Attribute Information: This is the 6x6 Scattering Matrix (S): S = [ s12, s13, s14, s15, s16, s21, s23, s24, s25, s26, s31, s32, s34, s35, s36, s41, s42, s43, s45, s46, s51, s52, s53, s54, s56, s61, s62, s63, s64, s65, ] The first 30 attributes (columns) of an instance are the 15 elements of the triangular upper part of S. Since each of these elements is a complex number with real and imaginary parts, each instance is a vector of 15x2=30 attributes. The real (real) and imaginary (img) parts of each element are placed one after the other. The scattering parameters are ordered row by row from left to right, e.g.: s12_real, s12_img, s13_real, s13_img, ..., s16_real, s16_img, s21_real, s21_img, ... . The self-scattering parameters, i.e. those placed on the main diagional of S, are not part of the dataset, as exaplined in [1]. Note: This dataset is realeased without pre-processing. For more information read the reference papers: [1] L. Urbinati, M. Ricci, G. Turvani, J. A. T. Vasquez, F. Vipiana and M. R. Casu, "A Machine-Learning Based Microwave Sensing Approach to Food Contaminant Detection," 2020 IEEE International Symposium on Circuits and Systems (ISCAS), Seville, Spain, 2020, pp. 1-5, doi: 10.1109/ISCAS45731.2020.9181293 Link to the paper: https://ieeexplore.ieee.org/abstract/document/9181293 [2] M. Ricci, B. Stitic, L. Urbinati, G. D. Guglielmo, J. A. T. Vasquez, L. P. Carloni, F. Vipiana, and M. R. Casu, "Machine-learning-based microwave sensing: A case study for the food industry," IEEE Journal on Emerging and Selected Topics in Circuits and Systems, vol. 11, no. 3, pp. 503-514, 2021 Link to the paper: https://ieeexplore.ieee.org/abstract/document/9489295

31 features

class (target)numeric2 unique values
0 missing
s12numeric1647 unique values
0 missing
s13numeric1397 unique values
0 missing
s14numeric1727 unique values
0 missing
s15numeric1658 unique values
0 missing
s16numeric1305 unique values
0 missing
s21numeric1521 unique values
0 missing
s23numeric1511 unique values
0 missing
s24numeric1254 unique values
0 missing
s25numeric1151 unique values
0 missing
s26numeric1198 unique values
0 missing
s31numeric1361 unique values
0 missing
s32numeric1127 unique values
0 missing
s34numeric535 unique values
0 missing
s35numeric898 unique values
0 missing
s36numeric1205 unique values
0 missing
s41numeric1256 unique values
0 missing
s42numeric970 unique values
0 missing
s43numeric920 unique values
0 missing
s45numeric585 unique values
0 missing
s46numeric756 unique values
0 missing
s51numeric902 unique values
0 missing
s52numeric845 unique values
0 missing
s53numeric1155 unique values
0 missing
s54numeric1201 unique values
0 missing
s56numeric799 unique values
0 missing
s61numeric723 unique values
0 missing
s62numeric696 unique values
0 missing
s63numeric467 unique values
0 missing
s64numeric860 unique values
0 missing
s65numeric1121 unique values
0 missing

19 properties

2400
Number of instances (rows) of the dataset.
31
Number of attributes (columns) of the dataset.
0
Number of distinct values of the target attribute (if it is nominal).
0
Number of missing values in the dataset.
0
Number of instances with at least one value missing.
31
Number of numeric attributes.
0
Number of nominal attributes.
0
Number of binary attributes.
0
Percentage of binary attributes.
0
Percentage of instances having missing values.
1
Average class difference between consecutive instances.
0
Percentage of missing values.
0.01
Number of attributes divided by the number of instances.
100
Percentage of numeric attributes.
Percentage of instances belonging to the most frequent class.
0
Percentage of nominal attributes.
Number of instances belonging to the most frequent class.
Percentage of instances belonging to the least frequent class.
Number of instances belonging to the least frequent class.

0 tasks

Define a new task