| Titre : | Realization of a Parallel Clustering method Application to agricultural soil management control |
| Auteurs : | OUMAIMA ANFEL BOUBAKEUR, Auteur ; Rachida Saouli, Directeur de thèse |
| Type de document : | Monographie imprimée |
| Editeur : | Biskra [Algérie] : Faculté des Sciences Exactes et des Sciences de la Nature et de la Vie, Université Mohamed Khider, 2020 |
| Format : | 1 vol. (72 p.) / ill. / 29 cm |
| Langues: | Anglais |
| Mots-clés: | Agricultural techniques,soil parameters,k-means algorithm,parallel calculation. |
| Résumé : | Agricultural techniques and treatments are usually applied on a plot of land although the depth and condition of the soil is not uniform. For this, we aim in this project, to use the clustering algorithm in order to divide the soil into groups of fields according to its parameters allowing farmers to have more precise information on the crop to use. Then, we propose a parallel technic based on k-means algorithm to provide an efficient technic that subdivide the soil into parts with different properties. In this case, we use a total and partial fusion technics leading to respectively the use of hierarchical or arbitrary merging between the clusters. To implement our split soil module we based on agricultural data base provided by the Direction of the Research Division "Soil Resource Management and Utilization in Arid Regions" of the CRSTRA (Centre for Scientific and Technical Research on Arid Regions) and we used MPI (Message passing Information) API on IBN KHALDOUN HPC (High performance Computer) of university of BISKRA. We show through the obtained results that the sequential implementation is very heavy time consuming compared to the parallel on 30 values per parameter in the used dataset. |
| Sommaire : |
General Introduction…………………………………………………. 1
Chapitre I I.1. Introduction ……………………………………………………… 3 I.2. Precision Agriculture……………………………………………. 3 I.2.1. Precision Agronomics…………………………………. 4 I.2.2. Categories of soils……………………………………… 5 A. Class 1 : Very good qualitysoils……………. 5 B. Class 2 : Good quality soils………………… 5 C. Class 3: Medium quality soils……………… 6 D. Class 4: Soils of poor quality……………….. 7 E. Class 5 :Poor to good quality soils…………. 8 F. Class 6 :Soils to be kept under natural cover… 8 I.3. Clustering methods in precision agriculture……………………. 9 I.3.1. Clustering……………………………………………….. 9 A. Clustering workflow……………………………. 9 1.Prepare data………………………………….. 9 2.Creat similarity metric………………………. 9 3.Run clustering algorithm……………………. 10 4.Interpret results and adjust your lustering…. 10 B.Clustering Methods………………………………. 10 B.1 Density-based Clustering…………………… 10 B.2. Distribution based Clustering……………… 13 B.3. Connectivity based Clustering……………… 15 B.4. Hierarchical Clustering…………………….. 15 B.4.1. Hierarchical agglomerative clustering 15 B.4.2. Hierarchical Divisive Clustering…….. 16 B.5. Centroid-based Clustering………………….. 17 I.4. Work synthesis………………………………………………………. 20 I.4.1. Fuzzy Clustering in plant disease detection……………… 20 I.4.2. Hierarchical Agglomerative Clustering in soil delineation. 21 I.5. Conclusion……………………………………………………………. 22 Chapitre II II.1 Introduction………………………………………………………….. 23 II.2 Modern Computing…………………………………………………. 23 II.2.1 Uniprocessor Computers…………………………………………. 23 II.2.2 Multiple Processors……………………………………………….. 25 II.3 Emergence of cluster computers……………………………………. 27 II.4 Cluster structure……………………………………………………... 28 II.5 High performance Network…………………………………………. 30 II.6 Conclusion……………………………………………………………. 34 Chapitre III III.1 Introduction…………………………………………………………. 35 III.2 Conception of Parallel k-means algorithm ………………………. 35 III.3 Split Soil Module……………………………………………………. 36 III.3.1. Sequential_ K-means_SS algorithm principle………. 36 III.3.2. Parallel K-means_SS algorithm principle…………… 37 III.3.3. FUSION……………………………………………….. 38 III.3.3.1 Total Fusion…………………………….. 38 III.3.3.2 Partial Fusion…………………………… 40 III.4 Implementation of parallel k-means algorithm……………….. 42 III.4.1 Description……………………………………………. 42 III.4.1.1 MPI……………………………………… 42 III.4.1.2 C Programming Language…………….. 42 III.4.1.3 HPC……………………………………… 43 III.4.1.4 CRSTRA………………………………… 45 III.4.2 Data description……………………………………….. 46 III.4.3. Parallel split soil………………………………………. 49 III.5 Results…………………………………………………………….. 55 III.5.1 Execution Process…………………………………. 55 III.5.2 Execution Results………………………………….. 58 III.5.2.1 Clustering Results……………………. 58 III.5.3 Fusion Results………………………………………… 62 III.5.4 Comparison…………………………………………….. 67 III.6 Conclusion………………………………………………... 68 Conclusion Coclusion……………………………………………………….. 69 |
| Type de document : | Mémoire master |
Disponibilité (1)
| Cote | Support | Localisation | Statut |
|---|---|---|---|
| MINF/540 | Mémoire master | bibliothèque sciences exactes | Consultable |
