تصوير العناقيد والرسم الشجري لوجوه الحيوانات - كتاب الذكاء الإصطناعي - الصف 12 - الفصل 1 - المملكة العربية السعودية

شكل 4.20: تصوير العناقيد

يُعدُّ هذا التصوير أكثر مصداقية من الذي تم إنتاجه للبيانات غير المحوّلة، وعلى الرغم من وجود بعض الشوائب، فإن الشكل يُظهر عناقيد واضحة ومفصولة جيدًا، ويمكن الآن حساب الرسم الشجري لمجموعة البيانات هذه.

plt.figure() linkage_2 = hierarchy.linkage(X_hog, method = 'ward') hierarchy.dendrogram(linkage_2) plt.show()

شكل 4.21: الرسم الشجري لفئات وجوه الحيوانات المختلفة باستخدام مخطط تكراري للتدرجات الموجهة (HOG)

وزارة التعليم Ministry of Education 2023 - 1447

228

شكل 4.20: تصوير العناقيد يُعدُّ هذا التصوير أكثر مصداقية من الذي تم إنتاجه للبيانات غير المحوّلة، وعلى الرغم من وجود بعض الشوائب، فإن الشكل يُظهر عناقيد واضحة ومفصولة جيدًا، ويمكن الآن حساب الرسم الشجري لمجموعة البيانات هذه.plt.figure() linkage_2 = hierarchy.linkage(X_hog, method = 'ward') hierarchy.dendrogram(linkage_2) plt.show()شكل 4.21: الرسم الشجري لفئات وجوه الحيوانات المختلفة باستخدام مخطط تكراري للتدرجات الموجهة (HOG)2023 - 1447--- VISUAL CONTEXT --- **GRAPH**: شكل 4.20: تصوير العناقيد Description: A 2D scatter plot titled 'TSNE Projection of 1085 Documents' showing clusters of data points. Each color represents a different animal category, as indicated by the legend. The clusters appear distinct and well-separated. The legend includes: Bear, Cat, Chicken, Deer, Duck, Eagle, Lion, Pigeon, Rabbit, Wolf. X-axis: Not explicitly labeled, represents a t-SNE dimension Y-axis: Not explicitly labeled, represents a t-SNE dimension Data: The graph displays 1085 data points projected into a 2D space, forming 10 distinct clusters. Each cluster is color-coded to represent a specific animal type (e.g., red for Chicken, green for Deer, blue for Bear). The points within each cluster are tightly grouped, and the clusters themselves are visually well-separated, indicating effective dimensionality reduction and clustering. Key Values: 1085 Documents, 10 animal categories: Bear, Cat, Chicken, Deer, Duck, Eagle, Lion, Pigeon, Rabbit, Wolf Context: This visual demonstrates the application of t-SNE for dimensionality reduction and data visualization, showing how complex data (documents related to animals) can be projected into a lower-dimensional space to reveal inherent clustering structures. It supports the text discussing the reliability of this visualization for non-transformed data.**DIAGRAM**: شكل 4.21: الرسم الشجري لفئات وجوه الحيوانات المختلفة باستخدام مخطط تكراري للتدرجات الموجهة (HOG) Description: A dendrogram illustrating the hierarchical clustering of different animal face categories. The y-axis represents dissimilarity or distance, ranging from 0 to 100. The x-axis implicitly represents individual data points or sub-clusters. The tree-like structure shows how smaller clusters merge into larger ones at increasing dissimilarity levels. Major clusters are highlighted with different colors (blue, green, red, yellow, grey). X-axis: Implicitly represents data points/clusters Y-axis: Dissimilarity/Distance Data: The dendrogram shows a clear hierarchical structure where data points (likely animal faces processed with HOG features) are progressively grouped. At low dissimilarity values, many small clusters are present. As the dissimilarity increases along the y-axis, these clusters merge, forming larger, more generalized groups. Distinct color-coded branches (blue, green, red, yellow, grey) indicate the formation of major clusters at certain dissimilarity thresholds. Key Values: Y-axis range: 0 to 100 (dissimilarity/distance) Context: This visual demonstrates hierarchical clustering, a method used to build a hierarchy of clusters. It shows how different animal face categories, processed using HOG features, are grouped based on their similarity, providing insights into the natural structure of the dataset. It directly relates to the preceding code block and text about calculating the dendrogram for the dataset.