Deep Learning Projects with PyTorch

This repository is a collection of Deep Learning application projects using PyTorch I have learned through series 'Deep Learning with PyTorch' provided by DataCamp.

The main goal is to strengthen fundamental knowledge of Deep Learning, get familiar with PyTorch and apply into real-world use cases. The projects cover diverse input types: structured data, sequences, images, text using multiple deep learning models (CNN, RNN, LSTM, GRU, Transformers) to tackle a variety tasks such as classification, prediction,...

Along the way, applying real-world projects, tinkering with tensors, dealing with preprocessing data and evaluating helps me to learn a lot. Check my mini project experiment on Substack blog for more detailed experiments.

---

Deep Learning basic and its application

Id	Data Types	Tag	Title	Description	Dataset	Methods/Models	Metrics	Results	Note
1	Structured data	Binary Classification	Detecting Cybersecurity Threats	Classify whether a threat or not	BETH dataset	Neural Networks	Accuracy	0.9448	Done
2	Image	Image Classification	Clouds image classification	Classify cloud images into category	Cloud image	Convolutional Neural Network	Precision, Recall		Done
3	Image	Multiclass Image Classification	E-commerce Clothing classifier	Classify clothing images into category	Clothing image	Convolutional Neural Network	Precision, Recall	0.87 0.77	Done
4	Sequence data	Time series prediction	Predicting electricity consumption		Electricity Consumption	RNN, LSTM, GRU	MSE, RMSE	0.04 0.2	Done
5	Sequence data	Time series prediction	Predicting traffic volume		Traffic volume by hour	RNN, LSTM, GRU	MSE, RMSE	0.071 0.26	Done
6	Text	Text Multiclass classification	Customer service text multiclass classification	Classify CS ticket into 5 categories	Customer Support ticket	CNN	Accuracy, Precision, Recall	0.7892 0.7931 0.7892	Done
7	Structured data	Regression	Concrete strength prediction	Predict concrete strength based on their attributes	Concrete Dataset	MLP	MSE, RMSE	136 11.68	Done
8	Text	Text classification	Sentiment Analysis using Fine-tuning BERT	Sentiment Analysis (Negative, Positive)		MLP	MSE, RMSE	136 11.68	Sample done

Advanced Deep Learning with fastai

Id	Lesson objectives	Tag	Applications	Note
-	-	-	-	-

---

References

• DataCamp PyTorch skill tracks
• Practical Deep Learning for Coders and Deep Learning Foundations to Stable Diffusion series

---

⚙️ Tech Stack

• PyTorch – Deep Learning framework
• Torchvision – Image datasets & transforms
• Scikit-learn – Preprocessing & evaluation metrics
• Matplotlib / Seaborn – Visualization
• Pandas & Numpy – Data manipulation