diff --git a/content/english/projects/project-4.md b/content/english/projects/project-4.md index 976c8c5..21720ab 100644 --- a/content/english/projects/project-4.md +++ b/content/english/projects/project-4.md @@ -1,17 +1,109 @@ --- -title: "Reasons Why Employees Quit | Human Resources Analytics" -date: 2020-03-14T15:40:24+06:00 -# Project thumb -image : "images/projects/project4.jpg" -draft: true -# description -description: "This is meta description" +title: "Classification of Musical Genres" +date: 2024-01-10T15:40:24+06:00 +image: "images/projects/music-classification.png" +draft: false +description: "A microservices architecture for automatic music genre classification using machine learning and deep learning approaches" --- -Lorem ipsum dolor sit amet, consetetur sadipscing elitr, sed diam nonumy eirmod tempor invidunt ut labore et dolore magna aliquyam erat, sed diam voluptua. At vero eos et accusam et justo duo dolores et ea rebum. Stet clita kasd gubergren, no sea takimata sanctus est Lorem ipsum dolor sit amet. Lorem ipsum dolor sit amet, consetetur sadipscing elitr, sed diam nonumy eirmod tempor +This project implements a comprehensive solution for automatic music genre classification through a microservices architecture. The system combines classical machine learning and deep learning approaches to achieve robust genre classification accuracy on audio inputs. -invidunt ut labore et dolore magna aliquyam erat, sed diam voluptua. At vero eos et accusam et justo duo dolores et ea rebum. Stet clita kasd gubergren, no sea takimata sanctus est Lorem ipsum dolor sit amet. Lorem ipsum dolor sit amet, consetetur sadipscing elitr, sed diam nonumy eirmod tempor invidunt ut labore et dolore magna aliquyam erat, sed diam voluptua. At vero eos et +#### System Architecture -accusam et justo duo dolores et ea rebum. Stet clita kasd gubergren, no sea takimata sanctus est Lorem ipsum dolor sit amet.Lorem ipsum dolor sit amet, consetetur sadipscing elitr, sed diam nonumy eirmod tempor invidunt ut labore et dolore magna aliquyam erat, sed diam voluptua. At vero eos et accusam et justo duo dolores et ea rebum. Stet clita kasd gubergren, no sea takimata +The application is built on a microservices architecture with three main components: -sanctus est Lorem ipsum dolor sit amet. Lorem ipsum dolor sit amet, consetetur sadipscing elitr, sed diam nonumy eirmod tempor invidunt ut labore et dolore magna aliquyam erat, sed diam voluptua. At vero eos et accusam et justo duo dolores et ea rebum. Stet clita kasd gubergren, no sea takimata sanctus est Lorem ipsum dolor sit amet. Lorem ipsum dolor sit amet, consetetur \ No newline at end of file +1. **Classification Service** + - Implements two parallel classification models: + - SVM classifier using extracted audio features (MFCC, spectral centroid, zero-crossing rate) + - VGG19 deep learning model trained on mel-spectrograms + - RESTful API endpoints for model inference + - Audio preprocessing pipeline for feature extraction + +2. **Frontend Service** + - Streamlit-based web interface + - Real-time audio file upload and processing + - Interactive visualization of classification results + - Model performance metrics display + +3. **Orchestration Layer** + - Docker containers for each service + - Jenkins pipeline for automated testing and deployment + - Load balancing and service discovery + - Azure cloud infrastructure management + +#### Technical Implementation + +**Audio Processing Pipeline** +- Audio file validation and format standardization +- Feature extraction using librosa: + - Mel-frequency cepstral coefficients (MFCCs) + - Spectral features (centroid, rolloff, bandwidth) + - Temporal features (zero-crossing rate, RMS energy) +- Mel-spectrogram generation for deep learning model + +**Machine Learning Models** +- Support Vector Machine (SVM): + - Kernel: RBF with optimized parameters + - Feature scaling and normalization + - Cross-validation for model evaluation + +- VGG19 Neural Network: + - Transfer learning from ImageNet weights + - Fine-tuning on GTZAN dataset + - Data augmentation techniques for robust training + - Batch normalization and dropout for regularization + +**Deployment Infrastructure** +- Containerized services with Docker: + - Base image optimization for reduced size + - Multi-stage builds for production deployment + - Volume mapping for model persistence + +- CI/CD Pipeline: + - Automated testing with pytest + - Code quality checks with SonarQube + - Automated Docker image building and pushing + - Blue-green deployment strategy + +#### Performance and Metrics + +- Model Accuracy: + - SVM: 78% accuracy on test set + - VGG19: 85% accuracy on test set + - Ensemble approach: 87% accuracy + +- System Performance: + - Average response time: <2 seconds + - Concurrent user support: up to 100 + - API throughput: 50 requests/second + +#### Development Workflow + +The project followed an agile development methodology: + +1. **Planning Phase** + - Architecture design and component specification + - Technology stack selection + - Development roadmap creation + +2. **Implementation Phase** + - Iterative development of services + - Regular integration testing + - Performance optimization + +3. **Deployment Phase** + - Infrastructure setup on Azure + - CI/CD pipeline configuration + - Production deployment and monitoring + +#### Future Improvements + +- Implementation of real-time audio classification +- Addition of more sophisticated ensemble methods +- Integration of user feedback for model improvement +- Expansion of supported audio formats +- Implementation of A/B testing framework + +[View Project Repository](https://github.com/yourusername/music-classification) + +[Live Demo](https://your-demo-url.com) \ No newline at end of file diff --git a/content/english/projects/project-5.md b/content/english/projects/project-5.md index a8fc604..8829299 100644 --- a/content/english/projects/project-5.md +++ b/content/english/projects/project-5.md @@ -3,7 +3,7 @@ title: "Reasons Why Employees Quit | Human Resources Analytics" date: 2020-03-14T15:40:24+06:00 # Project thumb image : "images/projects/project5.jpg" -draft: true +draft: false # description description: "This is meta description" --- diff --git a/public/index.xml b/public/index.xml index 76b9a16..c05eee0 100644 --- a/public/index.xml +++ b/public/index.xml @@ -44,6 +44,13 @@ http://localhost:1313/projects/project-2/ <p>BiblioManager is a library management application offering an intuitive user interface developed with JavaFX. It allows efficient management of book collections, borrowing and member memberships.</p> + + Classification of Musical Genres + http://localhost:1313/projects/project-4/ + Wed, 10 Jan 2024 15:40:24 +0600 + http://localhost:1313/projects/project-4/ + <p>This project implements a comprehensive solution for automatic music genre classification through a microservices architecture. The system combines classical machine learning and deep learning approaches to achieve robust genre classification accuracy on audio inputs.</p> + “The biggest change to News Feed since 2013”: Facebook is making it easy to bypass its algorithm Feed since biggest http://localhost:1313/publications/publication-1/ @@ -93,6 +100,13 @@ http://localhost:1313/talks/talk-4/ <p>Lorem ipsum dolor sit amet, consetetur sadipscing elitr, sed diam nonumy eirmod tempor invidunt ut labore et dolore magna aliquyam erat, sed diam voluptua. At vero eos et accusam et justo duo dolores et ea rebum. Stet clita kasd gubergren, no sea takimata sanctus est Lorem ipsum dolor sit amet. Lorem ipsum dolor sit amet, consetetur sadipscing elitr, sed diam nonumy eirmod tempor</p> + + Reasons Why Employees Quit | Human Resources Analytics + http://localhost:1313/projects/project-5/ + Sat, 14 Mar 2020 15:40:24 +0600 + http://localhost:1313/projects/project-5/ + <p>Lorem ipsum dolor sit amet, consetetur sadipscing elitr, sed diam nonumy eirmod tempor invidunt ut labore et dolore magna aliquyam erat, sed diam voluptua. At vero eos et accusam et justo duo dolores et ea rebum. Stet clita kasd gubergren, no sea takimata sanctus est Lorem ipsum dolor sit amet. Lorem ipsum dolor sit amet, consetetur sadipscing elitr, sed diam nonumy eirmod tempor</p> + Ahmed Sidi Mohamed http://localhost:1313/author/george-cushen/ diff --git a/public/projects/index.html b/public/projects/index.html index c4d8365..5a8358a 100644 --- a/public/projects/index.html +++ b/public/projects/index.html @@ -157,8 +157,38 @@

Maze Bank - +
+
+ +
+

Classification of Musical Genres

+

This project implements a comprehensive solution for automatic music genre classification through a microservices architecture. The system combines classical machine learning and deep learning approaches to achieve robust genre classification accuracy on audio inputs.

+
+
+
+
+
diff --git a/public/projects/index.xml b/public/projects/index.xml index e68583e..86852d9 100644 --- a/public/projects/index.xml +++ b/public/projects/index.xml @@ -30,5 +30,19 @@ http://localhost:1313/projects/project-2/ <p>BiblioManager is a library management application offering an intuitive user interface developed with JavaFX. It allows efficient management of book collections, borrowing and member memberships.</p> + + Classification of Musical Genres + http://localhost:1313/projects/project-4/ + Wed, 10 Jan 2024 15:40:24 +0600 + http://localhost:1313/projects/project-4/ + <p>This project implements a comprehensive solution for automatic music genre classification through a microservices architecture. The system combines classical machine learning and deep learning approaches to achieve robust genre classification accuracy on audio inputs.</p> + + + Reasons Why Employees Quit | Human Resources Analytics + http://localhost:1313/projects/project-5/ + Sat, 14 Mar 2020 15:40:24 +0600 + http://localhost:1313/projects/project-5/ + <p>Lorem ipsum dolor sit amet, consetetur sadipscing elitr, sed diam nonumy eirmod tempor invidunt ut labore et dolore magna aliquyam erat, sed diam voluptua. At vero eos et accusam et justo duo dolores et ea rebum. Stet clita kasd gubergren, no sea takimata sanctus est Lorem ipsum dolor sit amet. Lorem ipsum dolor sit amet, consetetur sadipscing elitr, sed diam nonumy eirmod tempor</p> + diff --git a/public/projects/page/2/index.html b/public/projects/page/2/index.html new file mode 100644 index 0000000..04c425b --- /dev/null +++ b/public/projects/page/2/index.html @@ -0,0 +1,222 @@ + + + + Projects + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + +
+ +
+ + +
+ +
+
+ + +
+
+ + +
+
+

Projects

+
+
+ + +
+
+
+ + +
+
+ +
+

Reasons Why Employees Quit | Human Resources Analytics

+

Lorem ipsum dolor sit amet, consetetur sadipscing elitr, sed diam nonumy eirmod tempor invidunt ut labore et dolore magna aliquyam erat, sed diam voluptua. At vero eos et accusam et justo duo dolores et ea rebum. Stet clita kasd gubergren, no sea takimata sanctus est Lorem ipsum dolor sit amet. Lorem ipsum dolor sit amet, consetetur sadipscing elitr, sed diam nonumy eirmod tempor

+
+
+
+ +
+ + +
+
+
+
+ +
+
+
+
+ +
+

+
+
+
+
+ + + + + + + + + + + + + + + + + + + + + + + + + \ No newline at end of file diff --git a/public/projects/project-4/index.html b/public/projects/project-4/index.html new file mode 100644 index 0000000..c64c3d5 --- /dev/null +++ b/public/projects/project-4/index.html @@ -0,0 +1,317 @@ + + + + Classification of Musical Genres + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + +
+ +
+ + +
+ +
+
+ + +
+
+ + +
+
+

Classification of Musical Genres

+
+
+ + +
+
+
+
+
+ +
    +
  • 24 Jan, 2024
    • +
  • read
    • + +
+ +
+
+

This project implements a comprehensive solution for automatic music genre classification through a microservices architecture. The system combines classical machine learning and deep learning approaches to achieve robust genre classification accuracy on audio inputs.

+

System Architecture

+

The application is built on a microservices architecture with three main components:

+
    +
  1. +

    Classification Service

    +
      +
    • Implements two parallel classification models: +
        +
      • SVM classifier using extracted audio features (MFCC, spectral centroid, zero-crossing rate)
        • +
      • VGG19 deep learning model trained on mel-spectrograms
        • +
      +
      • +
    • RESTful API endpoints for model inference
      • +
    • Audio preprocessing pipeline for feature extraction
      • +
    +
    2. +
  2. +

    Frontend Service

    +
      +
    • Streamlit-based web interface
      • +
    • Real-time audio file upload and processing
      • +
    • Interactive visualization of classification results
      • +
    • Model performance metrics display
      • +
    +
    3. +
  3. +

    Orchestration Layer

    +
      +
    • Docker containers for each service
      • +
    • Jenkins pipeline for automated testing and deployment
      • +
    • Load balancing and service discovery
      • +
    • Azure cloud infrastructure management
      • +
    +
    4. +
+

Technical Implementation

+

Audio Processing Pipeline

+
    +
  • Audio file validation and format standardization
    • +
  • Feature extraction using librosa: +
      +
    • Mel-frequency cepstral coefficients (MFCCs)
      • +
    • Spectral features (centroid, rolloff, bandwidth)
      • +
    • Temporal features (zero-crossing rate, RMS energy)
      • +
    +
    • +
  • Mel-spectrogram generation for deep learning model
    • +
+

Machine Learning Models

+
    +
  • +

    Support Vector Machine (SVM):

    +
      +
    • Kernel: RBF with optimized parameters
      • +
    • Feature scaling and normalization
      • +
    • Cross-validation for model evaluation
      • +
    +
    • +
  • +

    VGG19 Neural Network:

    +
      +
    • Transfer learning from ImageNet weights
      • +
    • Fine-tuning on GTZAN dataset
      • +
    • Data augmentation techniques for robust training
      • +
    • Batch normalization and dropout for regularization
      • +
    +
    • +
+

Deployment Infrastructure

+
    +
  • +

    Containerized services with Docker:

    +
      +
    • Base image optimization for reduced size
      • +
    • Multi-stage builds for production deployment
      • +
    • Volume mapping for model persistence
      • +
    +
    • +
  • +

    CI/CD Pipeline:

    +
      +
    • Automated testing with pytest
      • +
    • Code quality checks with SonarQube
      • +
    • Automated Docker image building and pushing
      • +
    • Blue-green deployment strategy
      • +
    +
    • +
+

Performance and Metrics

+
    +
  • +

    Model Accuracy:

    +
      +
    • SVM: 78% accuracy on test set
      • +
    • VGG19: 85% accuracy on test set
      • +
    • Ensemble approach: 87% accuracy
      • +
    +
    • +
  • +

    System Performance:

    +
      +
    • Average response time: <2 seconds
      • +
    • Concurrent user support: up to 100
      • +
    • API throughput: 50 requests/second
      • +
    +
    • +
+

Development Workflow

+

The project followed an agile development methodology:

+
    +
  1. +

    Planning Phase

    +
      +
    • Architecture design and component specification
      • +
    • Technology stack selection
      • +
    • Development roadmap creation
      • +
    +
    2. +
  2. +

    Implementation Phase

    +
      +
    • Iterative development of services
      • +
    • Regular integration testing
      • +
    • Performance optimization
      • +
    +
    3. +
  3. +

    Deployment Phase

    +
      +
    • Infrastructure setup on Azure
      • +
    • CI/CD pipeline configuration
      • +
    • Production deployment and monitoring
      • +
    +
    4. +
+

Future Improvements

+
    +
  • Implementation of real-time audio classification
    • +
  • Addition of more sophisticated ensemble methods
    • +
  • Integration of user feedback for model improvement
    • +
  • Expansion of supported audio formats
    • +
  • Implementation of A/B testing framework
    • +
+

View Project Repository

+

Live Demo

+ +
+
+
+
+
+ +
+
+
+
+ +
+

+
+
+
+
+ + + + + + + + + + + + + + + + + + + + + + + + + \ No newline at end of file diff --git a/public/projects/project-5/index.html b/public/projects/project-5/index.html new file mode 100644 index 0000000..c20d262 --- /dev/null +++ b/public/projects/project-5/index.html @@ -0,0 +1,175 @@ + + + + Reasons Why Employees Quit | Human Resources Analytics + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + +
+ +
+ + +
+ +
+
+ + +
+
+ + +
+
+

Reasons Why Employees Quit | Human Resources Analytics

+
+
+ + +
+
+
+
+
+ +
    +
  • 20 Mar, 2020
    • +
  • read
    • + +
+ +
+
+

Lorem ipsum dolor sit amet, consetetur sadipscing elitr, sed diam nonumy eirmod tempor invidunt ut labore et dolore magna aliquyam erat, sed diam voluptua. At vero eos et accusam et justo duo dolores et ea rebum. Stet clita kasd gubergren, no sea takimata sanctus est Lorem ipsum dolor sit amet. Lorem ipsum dolor sit amet, consetetur sadipscing elitr, sed diam nonumy eirmod tempor

+

invidunt ut labore et dolore magna aliquyam erat, sed diam voluptua. At vero eos et accusam et justo duo dolores et ea rebum. Stet clita kasd gubergren, no sea takimata sanctus est Lorem ipsum dolor sit amet. Lorem ipsum dolor sit amet, consetetur sadipscing elitr, sed diam nonumy eirmod tempor invidunt ut labore et dolore magna aliquyam erat, sed diam voluptua. At vero eos et

+

accusam et justo duo dolores et ea rebum. Stet clita kasd gubergren, no sea takimata sanctus est Lorem ipsum dolor sit amet.Lorem ipsum dolor sit amet, consetetur sadipscing elitr, sed diam nonumy eirmod tempor invidunt ut labore et dolore magna aliquyam erat, sed diam voluptua. At vero eos et accusam et justo duo dolores et ea rebum. Stet clita kasd gubergren, no sea takimata

+

sanctus est Lorem ipsum dolor sit amet. Lorem ipsum dolor sit amet, consetetur sadipscing elitr, sed diam nonumy eirmod tempor invidunt ut labore et dolore magna aliquyam erat, sed diam voluptua. At vero eos et accusam et justo duo dolores et ea rebum. Stet clita kasd gubergren, no sea takimata sanctus est Lorem ipsum dolor sit amet. Lorem ipsum dolor sit amet, consetetur

+ +
+
+
+
+
+ +
+
+
+
+ +
+

+
+
+
+
+ + + + + + + + + + + + + + + + + + + + + + + + + \ No newline at end of file