Changes

              
    

          
          
        
        

            
f
1
{
f
1
{
            
2
  "acknowledgement": "",
2
  "acknowledgement": "",
            
3
  "administrative_areas": "",
3
  "administrative_areas": "",
            
4
  "author": null,
4
  "author": null,
            
5
  "author_email": null,
5
  "author_email": null,
            
6
  "citation_narrative": "",
6
  "citation_narrative": "",
            
7
  "collaborators": "[{\"collaborator\": \"Fatiah Khamisi\"}]",
7
  "collaborators": "[{\"collaborator\": \"Fatiah Khamisi\"}]",
            
8
  "contact_person": "Daisy Salifu",
8
  "contact_person": "Daisy Salifu",
            
9
  "contact_person_email": "dsalifu@icipe.org",
9
  "contact_person_email": "dsalifu@icipe.org",
            
10
  "country": "[{\"country\": \"KE\"}]",
10
  "country": "[{\"country\": \"KE\"}]",
            
11
  "creator_user_id": "f09ec764-fe3c-4069-850b-f968ff0c20bb",
11
  "creator_user_id": "f09ec764-fe3c-4069-850b-f968ff0c20bb",
            
12
  "date_uploaded": "2021-10-17",
12
  "date_uploaded": "2021-10-17",
            
13
  "donor": "",
13
  "donor": "",
            
14
  "groups": [],
14
  "groups": [],
            
15
  "id": "4bd2ffd3-fc6c-4189-bce1-f54ca34ea011",
15
  "id": "4bd2ffd3-fc6c-4189-bce1-f54ca34ea011",
            
16
  "isopen": true,
16
  "isopen": true,
            
17
  "license_id": "cc-by",
17
  "license_id": "cc-by",
            
18
  "license_title": "Creative Commons Attribution",
18
  "license_title": "Creative Commons Attribution",
            
19
  "license_url": "http://www.opendefinition.org/licenses/cc-by",
19
  "license_url": "http://www.opendefinition.org/licenses/cc-by",
            
20
  "maintainer": "Kennedy Senagi",
20
  "maintainer": "Kennedy Senagi",
            
21
  "maintainer_email": "ksenagi@icipe.org",
21
  "maintainer_email": "ksenagi@icipe.org",
            
22
  "metadata_created": "2022-05-18T05:49:38.498118",
22
  "metadata_created": "2022-05-18T05:49:38.498118",
            
n
23
  "metadata_modified": "2022-05-18T05:55:37.980527",
n
23
  "metadata_modified": "2022-05-18T05:59:22.812673",
            
24
  "name": "african-fruit-fly-program",
24
  "name": "african-fruit-fly-program",
            
25
  "notes": "Taxonomic identity of fruitfly pest using morphometrics",
25
  "notes": "Taxonomic identity of fruitfly pest using morphometrics",
            
n
26
  "num_resources": 3,
n
26
  "num_resources": 4,
            
27
  "num_tags": 0,
27
  "num_tags": 0,
            
28
  "organization": {
28
  "organization": {
            
29
    "approval_status": "approved",
29
    "approval_status": "approved",
            
30
    "created": "2022-05-17T13:51:43.824504",
30
    "created": "2022-05-17T13:51:43.824504",
            
31
    "description": "",
31
    "description": "",
            
32
    "id": "b3f6f7ff-df08-49a0-8959-dc1444685c2f",
32
    "id": "b3f6f7ff-df08-49a0-8959-dc1444685c2f",
            
33
    "image_url": "2022-05-17-135143.809085plant.jpeg",
33
    "image_url": "2022-05-17-135143.809085plant.jpeg",
            
34
    "is_organization": true,
34
    "is_organization": true,
            
35
    "name": "plant-health",
35
    "name": "plant-health",
            
36
    "state": "active",
36
    "state": "active",
            
37
    "title": "Plant Health",
37
    "title": "Plant Health",
            
38
    "type": "organization"
38
    "type": "organization"
            
39
  },
39
  },
            
40
  "owner_org": "b3f6f7ff-df08-49a0-8959-dc1444685c2f",
40
  "owner_org": "b3f6f7ff-df08-49a0-8959-dc1444685c2f",
            
41
  "principal_investigator": "Daniel Masiga",
41
  "principal_investigator": "Daniel Masiga",
            
42
  "principal_investigator_email": "dmasiga@icipe.org",
42
  "principal_investigator_email": "dmasiga@icipe.org",
            
43
  "private": false,
43
  "private": false,
            
44
  "region": "East Africa",
44
  "region": "East Africa",
            
45
  "relationships_as_object": [],
45
  "relationships_as_object": [],
            
46
  "relationships_as_subject": [],
46
  "relationships_as_subject": [],
            
47
  "resources": [
47
  "resources": [
            
48
    {
48
    {
            
49
      "cache_last_updated": null,
49
      "cache_last_updated": null,
            
50
      "cache_url": null,
50
      "cache_url": null,
            
51
      "created": "2022-05-18T05:51:33.159984",
51
      "created": "2022-05-18T05:51:33.159984",
            
52
      "date_last_updated": "2022-05-18",
52
      "date_last_updated": "2022-05-18",
            
53
      "date_uploaded_resource": "2021-10-17",
53
      "date_uploaded_resource": "2021-10-17",
            
54
      "description": "",
54
      "description": "",
            
55
      "format": "XLSX",
55
      "format": "XLSX",
            
56
      "hash": "",
56
      "hash": "",
            
57
      "id": "0613d1d8-d0a2-4d34-9b4d-29f698855766",
57
      "id": "0613d1d8-d0a2-4d34-9b4d-29f698855766",
            
58
      "last_modified": "2022-05-18T05:51:33.141075",
58
      "last_modified": "2022-05-18T05:51:33.141075",
            
59
      "metadata_modified": "2022-05-18T05:51:33.155845",
59
      "metadata_modified": "2022-05-18T05:51:33.155845",
            
60
      "mimetype": 
60
      "mimetype": 
            
61
"application/vnd.openxmlformats-officedocument.spreadsheetml.sheet",
61
"application/vnd.openxmlformats-officedocument.spreadsheetml.sheet",
            
62
      "mimetype_inner": null,
62
      "mimetype_inner": null,
            
63
      "name": "Raw data, meta-data, data dictionary and protocol",
63
      "name": "Raw data, meta-data, data dictionary and protocol",
            
64
      "package_id": "4bd2ffd3-fc6c-4189-bce1-f54ca34ea011",
64
      "package_id": "4bd2ffd3-fc6c-4189-bce1-f54ca34ea011",
            
65
      "position": 0,
65
      "position": 0,
            
66
      "resource_type": null,
66
      "resource_type": null,
            
67
      "restricted": "{\"allowed_users\": \"\", \"level\": 
67
      "restricted": "{\"allowed_users\": \"\", \"level\": 
            
68
\"public\"}",
68
\"public\"}",
            
69
      "size": 734510,
69
      "size": 734510,
            
70
      "state": "active",
70
      "state": "active",
            
71
      "url": 
71
      "url": 
            
72
-29f698855766/download/fruitfly-morphometrics_data-repository-1.xlsx",
72
-29f698855766/download/fruitfly-morphometrics_data-repository-1.xlsx",
            
73
      "url_type": "upload"
73
      "url_type": "upload"
            
74
    },
74
    },
            
75
    {
75
    {
            
76
      "cache_last_updated": null,
76
      "cache_last_updated": null,
            
77
      "cache_url": null,
77
      "cache_url": null,
            
78
      "created": "2022-05-18T05:54:07.020692",
78
      "created": "2022-05-18T05:54:07.020692",
            
79
      "date_last_updated": "2022-05-18",
79
      "date_last_updated": "2022-05-18",
            
80
      "date_uploaded_resource": "2021-10-17",
80
      "date_uploaded_resource": "2021-10-17",
            
81
      "description": "",
81
      "description": "",
            
82
      "format": "HTML",
82
      "format": "HTML",
            
83
      "hash": "",
83
      "hash": "",
            
84
      "id": "e8e406c3-6816-4330-99b4-8ef68999eb76",
84
      "id": "e8e406c3-6816-4330-99b4-8ef68999eb76",
            
85
      "last_modified": null,
85
      "last_modified": null,
            
86
      "metadata_modified": "2022-05-18T05:54:07.015551",
86
      "metadata_modified": "2022-05-18T05:54:07.015551",
            
87
      "mimetype": null,
87
      "mimetype": null,
            
88
      "mimetype_inner": null,
88
      "mimetype_inner": null,
            
89
      "name": "[R codes] - Machine learning algorithms on insect 
89
      "name": "[R codes] - Machine learning algorithms on insect 
            
90
morphometrics data",
90
morphometrics data",
            
91
      "package_id": "4bd2ffd3-fc6c-4189-bce1-f54ca34ea011",
91
      "package_id": "4bd2ffd3-fc6c-4189-bce1-f54ca34ea011",
            
92
      "position": 1,
92
      "position": 1,
            
93
      "resource_type": null,
93
      "resource_type": null,
            
94
      "restricted": "{\"allowed_users\": \"\", \"level\": 
94
      "restricted": "{\"allowed_users\": \"\", \"level\": 
            
95
\"public\"}",
95
\"public\"}",
            
96
      "size": null,
96
      "size": null,
            
97
      "state": "active",
97
      "state": "active",
            
98
      "url": 
98
      "url": 
            
99
pe-official/Machine-Learning-Algorithms-on-Insect-Morphometrics-Data",
99
pe-official/Machine-Learning-Algorithms-on-Insect-Morphometrics-Data",
            
100
      "url_type": null
100
      "url_type": null
            
101
    },
101
    },
            
102
    {
102
    {
            
103
      "cache_last_updated": null,
103
      "cache_last_updated": null,
            
104
      "cache_url": null,
104
      "cache_url": null,
            
105
      "created": "2022-05-18T05:55:37.808750",
105
      "created": "2022-05-18T05:55:37.808750",
            
106
      "date_last_updated": "2022-05-18",
106
      "date_last_updated": "2022-05-18",
            
107
      "date_uploaded_resource": "2022-05-18",
107
      "date_uploaded_resource": "2022-05-18",
            
108
      "description": "Analysis of landmark-based morphometric 
108
      "description": "Analysis of landmark-based morphometric 
            
109
measurements taken on body parts of insects have been a useful 
109
measurements taken on body parts of insects have been a useful 
            
110
taxonomic approach alongside DNA barcoding in insect identification. 
110
taxonomic approach alongside DNA barcoding in insect identification. 
            
111
Statistical analysis of morphometrics have largely been dominated by 
111
Statistical analysis of morphometrics have largely been dominated by 
            
112
traditional methods and approaches such as principal component 
112
traditional methods and approaches such as principal component 
            
113
analysis (PCA), canonical variate analysis (CVA) and discriminant 
113
analysis (PCA), canonical variate analysis (CVA) and discriminant 
            
114
analysis (DA). However, advancement in computing power creates a 
114
analysis (DA). However, advancement in computing power creates a 
            
115
paradigm shift to apply modern tools such as machine learning. Herein, 
115
paradigm shift to apply modern tools such as machine learning. Herein, 
            
116
we assess the predictive performance of four machine learning 
116
we assess the predictive performance of four machine learning 
            
117
classifiers; K-nearest neighbor (KNN), random forest (RF), support 
117
classifiers; K-nearest neighbor (KNN), random forest (RF), support 
            
118
vector machine (the linear, polynomial and radial kernel SVMs) and 
118
vector machine (the linear, polynomial and radial kernel SVMs) and 
            
119
artificial neural network (ANNs) on fruit fly morphometrics that were 
119
artificial neural network (ANNs) on fruit fly morphometrics that were 
            
120
previously analysed using PCA and CVA. KNN and RF performed poorly 
120
previously analysed using PCA and CVA. KNN and RF performed poorly 
            
121
with overall model accuracy lower than \u201cno-information rate\u201d 
121
with overall model accuracy lower than \u201cno-information rate\u201d 
            
122
(NIR) (p value\u2009>\u20090.1). The SVM models had a predictive 
122
(NIR) (p value\u2009>\u20090.1). The SVM models had a predictive 
            
123
accuracy of\u2009>\u200995%, significantly higher than NIR 
123
accuracy of\u2009>\u200995%, significantly higher than NIR 
            
124
(p\u2009<\u20090.001), Kappa\u2009>\u20090.78 and area under curve 
124
(p\u2009<\u20090.001), Kappa\u2009>\u20090.78 and area under curve 
            
125
(AUC) of the receiver operating characteristics was\u2009>\u20090.91; 
125
(AUC) of the receiver operating characteristics was\u2009>\u20090.91; 
            
126
while ANN model had a predictive accuracy of 96%, significantly higher 
126
while ANN model had a predictive accuracy of 96%, significantly higher 
            
127
than NIR, Kappa of 0.83 and AUC was 0.98. Wing veins 2, 3, 8, 10, 14 
127
than NIR, Kappa of 0.83 and AUC was 0.98. Wing veins 2, 3, 8, 10, 14 
            
128
and tibia length were of higher importance than other variables based 
128
and tibia length were of higher importance than other variables based 
            
129
on both SVM and ANN models. We conclude that SVM and ANN models could 
129
on both SVM and ANN models. We conclude that SVM and ANN models could 
            
130
be used to discriminate fruit fly species based on wing vein and tibia 
130
be used to discriminate fruit fly species based on wing vein and tibia 
            
131
length measurements or any other morphologically similar pest taxa. 
131
length measurements or any other morphologically similar pest taxa. 
            
132
These algorithms could be used as candidates for developing an 
132
These algorithms could be used as candidates for developing an 
            
133
integrated and smart application software for insect discrimination 
133
integrated and smart application software for insect discrimination 
            
134
and identification. Variable importance analysis results in this study 
134
and identification. Variable importance analysis results in this study 
            
135
would be useful for future studies for deciding what must be 
135
would be useful for future studies for deciding what must be 
            
136
measured.",
136
measured.",
            
137
      "format": "HTML",
137
      "format": "HTML",
            
138
      "hash": "",
138
      "hash": "",
            
139
      "id": "cb7e4445-4440-4098-8b72-441a2597bb47",
139
      "id": "cb7e4445-4440-4098-8b72-441a2597bb47",
            
140
      "last_modified": null,
140
      "last_modified": null,
            
141
      "metadata_modified": "2022-05-18T05:55:37.803541",
141
      "metadata_modified": "2022-05-18T05:55:37.803541",
            
142
      "mimetype": null,
142
      "mimetype": null,
            
143
      "mimetype_inner": null,
143
      "mimetype_inner": null,
            
144
      "name": "[Article] - Leveraging machine learning tools and 
144
      "name": "[Article] - Leveraging machine learning tools and 
            
145
algorithms for analysis of fruit fly morphometrics",
145
algorithms for analysis of fruit fly morphometrics",
            
146
      "package_id": "4bd2ffd3-fc6c-4189-bce1-f54ca34ea011",
146
      "package_id": "4bd2ffd3-fc6c-4189-bce1-f54ca34ea011",
            
147
      "position": 2,
147
      "position": 2,
            
148
      "resource_type": null,
148
      "resource_type": null,
            
149
      "restricted": "{\"allowed_users\": \"\", \"level\": 
149
      "restricted": "{\"allowed_users\": \"\", \"level\": 
            
150
\"public\"}",
150
\"public\"}",
            
151
      "size": null,
151
      "size": null,
            
152
      "state": "active",
152
      "state": "active",
            
153
      "url": "",
153
      "url": "",
            
154
      "url_type": null
154
      "url_type": null
            
t
t
155
    },
            
156
    {
            
157
      "cache_last_updated": null,
            
158
      "cache_url": null,
            
159
      "created": "2022-05-18T05:59:22.824877",
            
160
      "date_last_updated": "2022-05-18",
            
161
      "date_uploaded_resource": "2021-10-17",
            
162
      "description": "",
            
163
      "format": "PDF",
            
164
      "hash": "",
            
165
      "id": "7524a0ed-79fc-47cf-a478-fb0b45988fe1",
            
166
      "last_modified": "2022-05-18T05:59:22.800928",
            
167
      "metadata_modified": "2022-05-18T05:59:22.819494",
            
168
      "mimetype": "application/pdf",
            
169
      "mimetype_inner": null,
            
170
      "name": "Protocol",
            
171
      "package_id": "4bd2ffd3-fc6c-4189-bce1-f54ca34ea011",
            
172
      "position": 3,
            
173
      "resource_type": null,
            
174
      "restricted": "{\"allowed_users\": \"\", \"level\": 
            
175
\"public\"}",
            
176
      "size": 456159,
            
177
      "state": "active",
            
178
      "url": 
            
179
/resource/7524a0ed-79fc-47cf-a478-fb0b45988fe1/download/protocol.pdf",
            
180
      "url_type": "upload"
            
155
    }
181
    }
            
156
  ],
182
  ],
            
157
  "state": "active",
183
  "state": "active",
            
158
  "tags": [],
184
  "tags": [],
            
159
  "third_party": "no",
185
  "third_party": "no",
            
160
  "title": "African fruit fly program",
186
  "title": "African fruit fly program",
            
161
  "type": "dataset",
187
  "type": "dataset",
            
162
  "url": null,
188
  "url": null,
            
163
  "version": null
189
  "version": null
            
164
}
190
}