rttrif/TrifonovRS.Deep_Learning_Portfolio.github.io

This portfolio was created in preparation for certification for the TensorFlow Developer

DEEP LEARNING PORTFOLIO

This portfolio was created in preparation for certification the TensorFlow Developer

---

👨‍🎓 MY CERTIFICATES IN THE FIELD OF DEEP LEARNING

TensorFlow Developer Certificate

[GeekBrains] Certificate of professional retraining under the program "Geek University Faculty of Artificial Intelligence"

• [Udemy] TensorFlow Developer Certificate in 2021: Zero to Mastery
• [Udemy] Hands-On Computer Vision with OpenCV 4, Keras & TensorFlow 2
• [Coursera] DeepLearning.AI TensorFlow Developer
• [GeekBrains] Introduction to Neural Networks
• [GeekBrains] Deep learning in computer vision
• [GeekBrains] Introduction to natural language processing
• [GeekBrains] Convolutional neural network, CNN

---

---

📈 REGRESSION

Regression: Project 1: Household Electric Power Consumption

🧩 CLASSIFICATION

Binary classification: Project 2: Heart Failure Prediction

Multiclass classification: Project 3: Fashion-MNIST

---

---

👁️ COMPUTER VISION

🧬 Design and training of basic architectures from scratch

CLASSIFICATION

---

1. Simple architectures CNN - Tiny VGG: Project 4: COVID-19 Radiography Database

2. LeNet-5: Project 5: Forest Fire Detection

3. AlexNet: Project 6: Vehicle Classification

4. VGG-16 (VGG-16 & VGG-19): Project 7: Animal Faces

5. ResNet (ResNet & ResNet-50): Project 8: Weather Classification

6. Inception V1: Project 9: Brain Hemorrhage

7. Xception: Project 10: Brain Tumor Classification

SEMANTIC SEGMENTATION

---

1. Autoencoder for semantic segmentation: Project 11: Underwater Segmentation

2. FCN: Fully Convolutional Network: Project 12: Cityscapes Segmentation

3. U-Net: Project 13: People Clothing Segmentation

4. DeepLab V3+: Project 14: Leaf disease segmentation

5. Mask R-CNN: Project 15: Car segmentation

TRANSFER LEARNING

---

1. Image classification: Project 16: The Food-101 Data Set

• Feature extraction: EfficientNet

• Fine-tuning: RegNetY

2. Semantic segmentation: Project 17: The Oxford-IIIT Pet Dataset

• Fine-tuning: U-net

3. Object detection: Project 18: YOLOv3 with Darknet Pre-trained Weights

• Pre-trained: Yolo v3

---

---

NATURAL LANGUAGE PROCESSING

1. Naive Bayes classifier with TF-IDF encoder: Project 19: NLP with Disaster Tweets

2. Feed-forward neural network: Project 20: Sms spam collection

3. Recurrent neural network: Project 21: Tweets Sentiment Classification

• Recurrent neural network: RNN

• Long short-term memory: LSTM

• Gated recurrent unit: GRU

• Bidirectional-LSTM

4. 1D Convolutional Neural Network: Project 22: What's Cooking?

5. Transfer Learning - Universal Sentence Encoder: Project 23: Toxic Comment Classification

6. Sequential sentence classification: Project 24: SkimLit

• TF-IDF classifier

• Conv1D with token embeddings

• Feature extraction with pretrained token embeddings

• Conv1D with character embeddings

• Combining pretrained token embeddings + character embeddings (hybrid embedding layer)

• Transfer Learning with pretrained token embeddings + character embeddings + positional embeddings

---

---

TIME SERIES, SEQUENCES AND PREDICTIONS

1. Naïve forecasting method: Project 31: Web Traffic Time Series Forecasting

2. Dense neural network: Project 32: Population Time Series Data

3. 1D Convolutional Neural Network: Project 33: Short-term electricity load forecasting

4. Long short-term memory: LSTM: Project 34: Amazon Stock Price 1997 to 2020

5. N-BEATS: Neural basis expansion analysis: Project 35: Bitcoin price forecasting

6. Project 36: Forecasting energy consumption for train traction (univariate time series)

7. Project 37: Forecasting energy consumption for train traction (multivariate time series)

---

---