Silabus 3 Hari Machine Learning Master Class
Durasi: 3 hari (09:00 – 16:00)
Istirahat: 12:00 – 13:00
Level: Master Class (Tingkat Lanjut)
Metode: Teori, Hands-on Coding, Studi Kasus, Diskusi
Hari 1: Fundamentals & Feature Engineering
Sesi 1: Introduction to Advanced Machine Learning (09:00 – 10:30)
- Tren terbaru dalam Machine Learning
- Perbedaan Traditional ML vs. Deep Learning
- Pipeline Machine Learning yang optimal
- Best practices dalam pengolahan dataset besar
Sesi 2: Feature Engineering & Data Preprocessing (10:30 – 12:00)
- Data preprocessing untuk model high-performance
- Teknik Feature Engineering: Scaling, Encoding, Transformasi
- Feature Selection & Dimensionality Reduction (PCA, LDA)
Istirahat (12:00 – 13:00)
Sesi 3: Model Selection & Hyperparameter Tuning (13:00 – 14:30)
- Pemilihan model optimal berdasarkan karakteristik data
- Grid Search vs. Random Search vs. Bayesian Optimization
- Tuning model untuk meningkatkan akurasi
Sesi 4: Model Evaluation & Interpretability (14:30 – 16:00)
- Evaluasi model dengan metrik yang tepat
- SHAP & LIME untuk interpretasi model
- Model deployment considerations
Hari 2: Deep Learning & Computer Vision
Sesi 1: Introduction to Deep Learning & Neural Networks (09:00 – 10:30)
- Arsitektur Neural Networks (CNN, RNN, Transformer)
- Optimasi model Deep Learning (Adam, RMSprop, SGD)
- Implementasi Feedforward Neural Network (Hands-on)
Sesi 2: Computer Vision with Convolutional Neural Networks (CNN) (10:30 – 12:00)
- Konsep CNN: Convolution, Pooling, Fully Connected Layer
- Implementasi CNN untuk klasifikasi gambar
- Transfer Learning dengan ResNet, EfficientNet
Istirahat (12:00 – 13:00)
Sesi 3: Advanced Deep Learning Techniques (13:00 – 14:30)
- Attention Mechanism & Transformer Model (Vision Transformer, ViT)
- Generative Adversarial Networks (GANs) untuk Synthetic Data
- Teknik Fine-Tuning untuk Custom Dataset
Sesi 4: Object Detection & Image Segmentation (14:30 – 16:00)
- Implementasi YOLO, SSD, dan Faster R-CNN
- Image Segmentation dengan U-Net dan Mask R-CNN
- Hands-on: Object Detection pada dataset real
Hari 3: NLP, Time-Series Forecasting & Deployment
Sesi 1: Natural Language Processing (NLP) & Transformers (09:00 – 10:30)
- Teknik NLP modern (Word Embeddings, Transformers, BERT, GPT)
- Sentiment Analysis & Text Classification
- Hands-on: Implementasi Transformer Model
Sesi 2: Time-Series Forecasting & Anomaly Detection (10:30 – 12:00)
- Teknik forecasting dengan LSTM dan ARIMA
- Anomaly detection menggunakan Autoencoders
- Hands-on: Prediksi data time-series
Istirahat (12:00 – 13:00)
Sesi 3: Model Deployment & MLOps (13:00 – 14:30)
- Model deployment dengan Flask & FastAPI
- Model monitoring dan CI/CD dalam MLOps
- Hands-on: Deployment model ke cloud
Sesi 4: Case Study & Capstone Project (14:30 – 16:00)
- Studi kasus real-world dari berbagai industri
- Implementasi end-to-end project
- Presentasi hasil & evaluasi
Output dari Master Class:
✅ Pemahaman mendalam tentang Machine Learning & Deep Learning
✅ Hands-on proyek dengan dataset real
✅ Teknik optimalisasi model & deployment best practices
✅ Studi kasus industri untuk penerapan nyata
