Initial AI exploration for the backend of the applications #50
Conversation
|
![github-advanced-security[bot]](https://avatars.githubusercontent.com/in/57789?s=60&v=4){kind=small}
| def evaluate_clustering(embeddings, n_clusters=5): | ||
| kmeans = KMeans(n_clusters=n_clusters, random_state=42) | ||
| labels = kmeans.fit_predict(embeddings) | ||
| silhouette_avg = silhouette_score(embeddings, labels) |
Check notice
Code scanning / SonarCloud
Results that depend on random number generation should be reproducible Low
|
|
||
| def visualize_embeddings(embeddings, labels, method='PCA'): | ||
| if method == 'PCA': | ||
| reducer = PCA(n_components=2) |
Check notice
Code scanning / SonarCloud
Results that depend on random number generation should be reproducible Low
| if method == 'PCA': | ||
| reducer = PCA(n_components=2) | ||
| elif method == 't-SNE': | ||
| reducer = TSNE(n_components=2, perplexity=30, n_iter=1000) |
Check notice
Code scanning / SonarCloud
Results that depend on random number generation should be reproducible Low
|
|
|
||
| # Choose dimensionality reduction method | ||
| if method == 'PCA': | ||
| reducer = PCA(n_components=2) |
Check notice
Code scanning / SonarCloud
Results that depend on random number generation should be reproducible Low
| if method == 'PCA': | ||
| reducer = PCA(n_components=2) | ||
| elif method == 't-SNE': | ||
| reducer = TSNE(n_components=2, perplexity=30, n_iter=1000) |
Check notice
Code scanning / SonarCloud
Results that depend on random number generation should be reproducible Low
|
|
||
| # Choose dimensionality reduction method | ||
| if method == 'PCA': | ||
| reducer = PCA(n_components=2) |
Check notice
Code scanning / SonarCloud
Results that depend on random number generation should be reproducible Low
| if method == 'PCA': | ||
| reducer = PCA(n_components=2) | ||
| elif method == 't-SNE': | ||
| reducer = TSNE(n_components=2, perplexity=30, n_iter=1000) |
Check notice
Code scanning / SonarCloud
Results that depend on random number generation should be reproducible Low
This pull request includes several changes to the data files and a new Jupyter notebook. The most important changes include the addition of new program entries in
Program.csv, new program types inProgramType.csv, and the creation of an empty Jupyter notebook file.Updates to data files:
notebook/data/Program.csv: Added multiple new program entries with details such as code, credits, title, URL, cycle, and ID.notebook/data/ProgramType.csv: Added new program types including Maîtrise with project, Microprogramme, Maîtrise with mémoire, Concentration in technologies of health, Doctorat, Programme court, Certificat, DESS, Cheminement universitaire en technologie (CUT), and Baccalauréat.New Jupyter notebook:
notebook/gnn.ipynb: Created a new empty Jupyter notebook with Python as the specified language.